Classical music as enforced Utopia

In classical music composition, whatever thematic or harmonic conflicts may be engineered along the way, everything always turns out for the best. Similar utopian thinking underlies performance: performers see their job as faithfully carrying out their master’s (the composer’s) wishes. The more perfectly they represent them, the happier the result. But why should performers not have a critical role to play in re-presenting a score, just as actors are permitted – required even – to find new meanings and new relevance in texts? And what or whom are performers obeying, the long dead composer (and what is the ethical basis for that?) or a policing system (teachers, examiners, adjudicators, critics, agents, promoters, record producers) that enforces an imaginary tradition from childhood to grave? Starting from the evidence of early recordings, showing that composers are misrepresented, this article seeks to unpick some of the delusions that support classical music practice.


Some delusions
It seems increasingly obvious that western classical music is essentially an oral culture imagining itself to be a written culture. Musicians learn their craft via practical, one-to-one studio teaching (supplemented by workshops, masterclasses and rehearsals) in which beliefs about how scores should be played are passed on orally and by example, sharing the assumption that the notation already encodes most of the information one needs from the past about how the music should sound. 'Historically informed performance' (HIP) sometimes refers to written documentation (especially performance treatises) from the past; but after nearly half a century of modern practice HIP is now substantially passed on as a collection of agreements among HIP players about what is stylish (Leech-Wilkinson and Prior, 2014: 50-52). The earliest recordings (from ca. 1900) offer us the first chance to compare treatises with the playing of their authors: the comparison suggests that written information (which is generally far more detailed ca. 1900 than in the past) is far from sufficient to allow performers to reconstruct the sounds we hear the authors of these treatises make on record (Leech-Wilkinson, 2012: para. 3.4). How much truer must that be of the more distant past? Recordings show that performance changes very greatly over time, that we cannot know how it sounded much before the 1870s (the early maturity of the oldest performers on record), and that scores do not encode nearly as much information from the past as we think they do. However, you look at it, classical music practice is an oral tradition to which writing contributes much less than we maintain.
A second delusion, one deeply embedded in teaching, practice, criticism and even legislation (which ascribes rights in a score to composers but not to performers who turn it into music), is the notion that composers are the creatives and performers their faithful servants, reproducing their intentions in sound, performing beautifully yet dutifully, as a lovely spouse carrying out her domestic duty to perfection. Musicians do perform wonderfully at the moment, perhaps more perfectly than ever before. But that does not mean that all is well, or that we are not dealing with a master-slave relationship or, simply, a delusion. Again, 120 years of recordings provide the definitive evidence: as performance styles change over time, our understandings of compositions and of their composers change too (Leech-Wilkinson, 2013). Performers, in other words, have been doing a very large part of the meaning making all along. And, far from seeing this as unwarranted interference that we should try to tame, we would do better to recognise it as a truth of musical practice, one to be celebrated, and one for which performers are owed recognition and reward.
I shall illustrate this briefly next, but what I really want to ask is what follows. Because although it seems inconceivable from within classical music's ideological bubble, we can see from both recordings and experiment that scores could be performed wonderfully quite differently, and there seems every reason, if we want classical music to find new audiences and to have a healthy future, to look for quite new ways of making the same notes work.

Historical evidence
A video recording of Arturo Benedetti Michelangeli playing Brahms's Ballade Op. 10 No. 4 in b minor shows us much about the way Brahms is currently constructed. 1 Michelangeli spends a full 14 s, after raising his hands to the keyboard, preparing himself mentally for the great challenge of enabling this masterpiece to speak. The performance is dreamy, his expression reverential. We worship the composer for his exquisite creation.
Ilona Eibenschu¨tz recorded the same score within a series of 'Reminiscences of Brahms' recorded in 1952. 2 Eibenschu¨tz knew Brahms. She was one of Clara Schumann's pupils and was the first person to whom he played his Op. 118 or 119 piano pieces: she reports overhearing him saying 'She is the pianist I best like to hear playing my works' (Evans, 2012). She recorded in 1903She recorded in and again in 1951She recorded in -1952, and the piece she recorded on both occasions (the Ballade in g minor Op. 118 No. 3) is performed almost identically. So there are good reasons to trust this 1952 performance of Op. 10 No. 4 as passing on some historical information about how Brahms sounded within his own circle. Her speed is 60 per cent faster than Michelangeli's (concordant with the generally faster speeds of her generation compared to recent, more reverential performances): the piece comes across as light, almost salon music, certainly no object of worship. Nor is there any sense in her reminiscences that she thought of Brahms as divine. But this is what happens to composers once they have been gone a while. They become prophets at least, passing on the Word which the performer's job is to convey faithfully to the people.
There is much more evidence, gathered together in the remarkable study of Scott (2014), that Eibenschu¨tz's recordings, and those of other pupils of Clara Schumann, pass on Brahms's intentions relatively faithfully (also Musgrave, 2003). Their performances are radically unlike ours, much less predictable, with rubato used to an extent that to us is incomprehensible as 'musical', rushing ahead with crescendi, slowing with diminuendi, dislocating the hands, speeding across phrase-ends rather than slowing down. None of this is allowable today. So how faithfully are Brahms's wishes being carried out now? And if a pianist were to give us the historical Brahms today -light, uneven, irregular, unpredictable -which exams could they pass, who would promote them, who would hire them to give a high-profile concert or make a recording?
Carl Reinecke's playing of Mozart in 1905 is in some ways even further from modern practice (and just as far from modern fortepiano playing, 'historically informed' by written documents and early pianos). 3 A leading nineteenth-century conductor, composer and teacher, Reinecke, was born in 1824, the year of Beethoven's ninth symphony. One might think his view of Mozart would be of some interest to musicians concerned (as all claim to be) to realize the composer's intentions, if not relevant to Mozart (who died, after all, in 1791) then to music of the mid-nineteenth century (Schumann, say) when Reinecke's own performance style would have matured. Reinecke breaks just about every rule of modern piano playing: his hands are almost never synchronized, he spreads chords, he uses marked rhythmic inequality during melodic flourishes, and much more; and so his evidence is simply not acceptable within the current model of faithfulness to a composer.
There are hundreds, perhaps thousands of examples like this from the earliest years of recordings, offering insights into nineteenth-century practices which sound bewilderingly unlike our own (Brown, 1999;Cook, 2013;Leech-Wilkinson, 2009a;Milsom, 2003;Peres da Costa, 2012;Philip, 1992). Music has changed more than one could ever have imagined. Scores we think must be performed one way were performed quite differently a century ago. For composers then alive or recently dead that means that we are unquestionably not being faithful to composers' wishes in modern performances. And it shows too, and this is much more important, that the same notes can make (and have made) substantially different music.
And yet, the whole ethos and structure, the entire system that constitutes music teaching and examining and writing about musical performance today, from first lessons, through graded examinations, the conservatories, the criteria by which people are judged as musical or unmusical players, as suitable or unsuitable to give concerts and make recordings, the judgements of agents, producers, critics and connoisseurs, all depend on the belief that we know best and agree on how these pieces are supposed to go, and that if they do not go that way we are not hearing the music of their composers and not behaving as competent musicians.
Yet clearly we are not hearing the music of their composers when we do it our way, and equally clearly competence can only be judged in relation to a particular performance style. So the notion that music teachers, examiners, critics, agents and the rest know how scores ought to sound is a delusion. They know only what they think is proper at the moment. To justify that preference by calls on the composer, and on claims about what was intended in the past, requires a level of knowledge of the performing past before ca. 1900 that no one has or will ever have. Our knowledge of the performance past is inadequate, therefore, to justify the enforcement of performance norms. Historically, not just theoretically, the claim that we know what is appropriate to a composer or a composition from much before 1900 is bogus. And for it to be true for music after 1900 we shall have to pay a lot more attention to period recordings than we do at present.

Ethics
Is the claim also ethically unfounded? Suppose we did know how a score used to go, as we may for late Brahms and as we do for Grieg, Debussy and every composer thereafter: what obligation do we have to reproduce that? A worn out question, you might think; for so much has been written by performers, musicologists and music philosophers about the performer's duty to the composer. 4 But it is a question that demands, now we can hear the size of the problem, to be readdressed. As ever with ethics, the underlying question is who is harmed? Who is harmed when the Brahms ballade, Op. 10 No. 4, is played by Michelangeli rather than Eibenschu¨tz? Certainly not Brahms. We imagine him harmed because we imagine him present, listening in as we ask a pupil, 'What does Brahms want in b. 23?' (Note the present tense.) But Brahms is dead. His estate is long-since wound up and there are no lingering issues of intellectual property. Or if you think there are then you have a very long list of performers and teachers, examiners and critics to start suing.
Is the listener harmed by hearing a different reading of a text? If they are then what damage is done to them when they go to the theatre to see Shakespeare? Do we feel damaged by seeing King Lear in a new light? On the contrary, that is why we go. Yet for classical music, the very idea of changing a score's meaning seems scandalous.
A photograph in the archive of the Royal Shakespeare Company of a moment from a 2009 performance of Hamlet shows the Hamlet (David Tennant), dressed in dinner jacket, holding up a modern barber's mirror. 5 The production is set in the claustrophobic, illusion-peddling world of Euro royals: their palaces, their taste, their money, their emotional and moral confusion. Nothing unusual there, you may think: this is a modern production of an early seventeenth-century text which vividly shows its relevance to our times, imagining how easily such a story might find its way into the pages of magazines like Hello! Shakespeare's text was written around 1600, published in 1603, contemporary with John Dowland's second and third Bookes of Songs and William Byrd's Gradualia. Why then would we not give those texts performances, set in other period sound worlds, that said something pertinent about their continuing relevance to us? Why not Byrd as political protest? Such things might occasionally happen, but never as an everyday performance in a typical concert setting. There they would be regarded as populist and perverted, a travesty of the composer's intentions (as if an Oxbridge choir is not). What is our problem with innovation, or even imagination, in musical performance? Why are we so horrified at giving classical music the same licence we give classical theatre?

Utopian dystopias
Perhaps because we practise classical music as Utopia, a perfect society, walled off from the rest of the world, which it would be unforgivable to disrupt. Utopias work when everyone follows the same rules. The rules seem good, they have settled into place through long practice and are known to work, and we agree to abide by them. But to do that requires authority: rules entail a ruler. The ruler for these purposes is imagined as the composer. But the composer is just a figurehead, a way of packaging up all our supposed obligations into one, the composer's intentions, and of thereby protecting that package from any hint of criticism. For Utopias are inherently totalitarian: everyone has to obey the beloved leader.
Upbringing makes these rules normative. At first, this is not a problem. The young musician accepts their teacher's word, learns the rules: play what the composer says (sic.), get praise. Young musicians feel they are being creative as they learn the moves their body needs to make to sound acceptably expressive: as the moves start to work they are happy to accept the beloved leader as the source of their delight. Normativity constructs nature through practice (in this case literally practising one's instrument). Or you could think of it as naturalisation supported by false history ('this is Mozartian, this is Brahmsian'). Either way, as the years of training accumulate, the word of one's teacher, the examples of one's idols, the requirements of examinations and competitions fuse with one's own experience of thousands of hours of practice, affording increasingly expert results, to embed within one's body the conviction that these musical behaviours are natural. One seems to be drawing on laws of musical nature, finding their ideal expression in one's playing and reaffirming it in every successful performance. Norms become naturalized.
But norms are nonetheless oppressive, doubly so when based on false history; and as the student learns to behave within them (accepting 'heavy community censorship', Hill, 2012: 89) they are also aware of gradually increasing fear: fear of making a mistake, fear of playing out of style, fear of non-conforming, of being judged unsuitable for work, of being judged 'unmusical' -that ultimate judgement which Kingsbury (1988: 65), writing of conservatoire culture, calls 'a statement of doom'. 'Musical', at conservatoire, becomes the stamp of approval on your Party membership card: if you have it you are allowed to seek work, if not, not. With fear comes stress, anxiety, and performance-related illness, a plague now for which the ideology may well be substantially responsible (Ginsborg et al., 2012).
The performance police are everywhere. Teachers, examiners, adjudicators, agents, critics, promoters, producers, record reviewers, bloggers. Performance is policed from first lesson to farewell recital. Above all, and right through their lives, performers police themselves, for this is a classic Gramscian hegemony. By subscribing to the elite values of their professional models, performers sustain and nurture their enslavement, and that of their young students, to this constructed world of artificial performance norms.
But like all police states, this one too is based on lies. The first is that musical works exist. This is a remarkably persistent meme which became embedded during the nineteenth century (Goehr, 1992) and is now so normative that it is hard to see how false it is. Scores exist, but music does not; rather it happens (Leech-Wilkinson, 2012). That it happens similarly every time a score is performed reinforces the sense that something stable must lie beneath. But in fact, our sense of a work rests in nothing more than recollections of moments of previous performances. When we speak of 'Beethoven's fifth', we seem to be speaking of something (some thing), but in fact we are speaking of a potential experience over time, or of some aspect of that experience, or of the composition that preceded the score, or of the score that transmits what little information survives concerning what the composer imagined. There is nothing 'out there', nothing existing, except the score. The danger with thinking of music in terms of these notional works is that one then easily comes to suppose that they have a fixed form which it is one's job as a performer to find once again.
The second sustaining lie is that there are or were composers' intentions, that they are known, and that they matter. Composers often, perhaps usually, imagine music in great detail as they compose, and they write down some aspect of what they imagine into a score. But they are often very happy for performers to show them ways they had not imagined of sounding their scores. Often they work with the performance style of particular performers in mind; always they write with the general performance style of their place and time as the environment in which they imagine music. There is no stepping outside the performance stylistic world in which one lives (Leech-Wilkinson, 2009b). Composers' intentions, then, are limited by when and where they live; they are not communicable to performers except in a very sketchy way, supplemented, while they are alive and available to be asked, by whatever they can explain in words or by example; their intentions can be modified or willingly set aside by things performers do that they find they like; and when they are dead they have none at all any more. They certainly cannot encompass all the possibilities afforded by their scores.
For the third lie is that there are limited possibilities inherent in a score, perhaps only one ideal performance. Recordings make clear that this is simply not true. There have already been hugely different musical characters found in scores that have been recorded through the past century. We have seen some examples already. Moreover, experiment with performers quickly shows that scores can be persuasively performed (that is to say, make convincing musical sense to one who knows nothing of the score or its recent performance tradition) in ways that are as far as possible from the indications contained in the score. In a previous article, I provided the example of Beethoven's 'Moonlight' sonata, which makes an excellent storm in the first movement played allegro furioso, a touching Mendelssohnian slow movement in the second played andante, and so on (Leech-Wilkinson, 2013: 51-52). That this was not Beethoven's intention is clear, but that does not prevent these approaches working very well and entirely unproblematically to anyone who has never heard the piece played another way. Refuting such performances, in other words, has to be done on ideological grounds, not musical.
One could add the lie that someone else -teachers, but ultimately the beloved leader -always knows best. Why are performers so subservient, so obedient? In another study I discuss this in relation to infant/carer vocalization and its relationship with adult music. Here it will be more helpful to point back to the education process through which musicians have passed. It is impossible to qualify as a performer without accepting current performance norms: the system selects only those who are obedient; the rest give up music along the way. And so careers are denied to those who do not care to conform. This is the nature of Utopia: you conform or you leave.

Creativity, real and imagined
Where, then, is the creativity which classical musicians feel is part of their everyday engagement with composers' scores? First it lies in the small variations that are acceptable within everyday practice. How extensive is the range within which performers are licensed to vary performances itself varies from generation to generation. Recordings show us that in the early twentieth century there were many different approaches still to be heard at one and the same time. D'Albert, de Pachmann, Cortot, Godowsky, were very different kinds of musician. National styles were still audibly different then. By contrast, between the 1950s and 2000s, the range was rather narrow. Over the past decade or two, greater variety has begun to reappear among younger musicians; and it is partly this that leads me to believe that the time to address these problems head on, with some hope of change as a result, may have come. So there is creativity, and performers feel that they exercise it, but most of the differences between performances that result are unnoticeable by non-specialists. This is curious. And one wonders why non-specialist audiences continue to attend concerts knowing exactly how each piece will seem to them. Partly this must be a matter of the comfort of the known. But then again, audiences for conventional concerts are ageing. So there does seem to be a problem.
If there is some creativity for performers, what about innovation? It seems that those two concepts (variation within, as opposed to changes to, performance norms) need to be separated in thinking about the performance practice of classical music today. Innovation, as we have seen in thinking about Byrd, is clearly not allowed in approved practice. And yet it does happen, albeit by accident and in tiny steps, too small for anyone to observe. This is the conclusion of studies into performance style change (see Leech-Wilkinson, 2009b). Performance style does change, massively over time, and that process must involve a sequence of innovative steps away from norms. That we never hear these moments of change individually (if we did the innovators would be condemned) shows that each must be tiny, but they nonetheless must be picked up unawares by other performers, and thus gradually spread around in a process of gradual cultural evolution (Leech-Wilkinson, 2009b: 256-257).
Why does change have to happen by accident? Given that, as we can now see, it is certain to happen anyway, and has been happening for a century past, and presumably continuously before then, why cannot we accept it as natural -genuinely so since it is a process of natural selection of the most useful accidental mutations -and then drive it quite deliberately ourselves, creating a culture of innovation with all the critical and economic benefits that could bring? One only has to look back to the invention of HIP in the late 1960s, and its continuous evolution since, to see how over time a forcibly innovative practice can become acceptable as a parallel stream of performance practice, with its own departments in conservatories, increasingly assimilated into the mainstream to the point where even experts can find it difficult to be sure whether historical or modern instruments are being used (Kenyon, 2013).
Why, then, do we need a police state for classical music? Why are we so frightened of alternative practice? We have already seen how powerful are the effects of naturalization through practice and repetition. I pursue elsewhere (in a study in progress) the possibility that the comfort people find in music contributes to its tendency to infantilise participants. Add in a teaching system in which nanny knows best, with teachers falling back on that oldest trick of poor parenting, 'Play like this because I say so', refined by the killer, 'If you don't you won't get work.' And you have a system in which musicians feel that there is no alternative to doing as they are told.
It is highly efficient, of course. It makes the rehearsals run on time. For there is also a powerful economic imperative enforced by employers: the clearer everyone is about how each score is to be played, the less rehearsal is needed, the cheaper the costs, the smaller the loss. Yet repetition palls and audiences decline: and so the imperative to cut costs becomes ever harsher. Not only, then, are musicians slaves to a fantasy of the omnipotent composer, but at the same time they find themselves subject also to the all-too-real demands of promoters. And the former prevents the latter from exploring a perfectly obvious capitalistic solution: to innovate (at the cost of a modicum of R&D investment) so as to attract new audiences generating profits to fund the investment. Perhaps it is time on economic grounds, then, as well as historical, ethical and artistic, to consider the possible benefits of innovation. For the good of musicians' psychological health, there is a strong case to be made for allowing greater agency through personal creativity. In sum, there are a plenty of reasons why musicians might benefit from facing down the performance police, taking responsibility for their own performances, and discovering new kinds of musicalities through which to allow these endlessly reproduced scores to mean something new.
It is not easy. Clearly our sense of what is musical cannot lie outside the musicality that constructs us, because we have to believe in order to perform persuasively. Performance styles have to become embodied, and that takes time, experiment and practice. But we cannot begin without challenging those expressive and interpretative habits that we have allowed to construct us as musicians. We can subvert them by constructing alternative performance styles, while recognising that subversion has to be a continuous process of change, not the establishment of new norms. That was HIP's mistake: that it attempted to find the correct way to perform a score, measured against an imagined historical original, and then to enforce it forever. But although their ideology was sick, HIP musicians were fantastically inventive in making a new style. We need to be that inventive again, only this time not to pretend that we are either historical or correct.
What is not ethically acceptable, I suggest, is to recognise the problem in theory and then do nothing practical about it: there has to a practical project alongside this dismantling of the beliefs underpinning classical music's utopian police state. It has to offer examples of how innovation can work, that is to say, how performances can be made which are not enslaved to current beliefs about the styles proper to composers or scores, nor to the scores themselves (bearing in mind that the composers are dead and unable to be distressed to find that their scores can be powerfully performed in ways quite other than they imagined). Just as with HIP, only thrilling examples of new approaches to old scores will persuade musicians to take the enormous risk of changing their beliefs and practices.

Radical performance
So we need a radical performance project, generating very different performances, for which I propose nine aims: 4. To test the possibilities for and consequences of freeing classical performance from its self-imposed framework of moral obligations 5. To follow the logic of the process of performance style change in order to see where else it might go 6. To explore the extent to which the notions 'musical' and 'unmusical' are ideologically grounded 7. To give performers due credit for the huge contribution they make to the creation of music, and correspondingly to reduce the role of composer to approximate that of playwright 8. To enrich our concert life by introducing the notion of production into performance of scores, with commercial benefits for artists, promoters, record companies, and attracting new audiences 9. To open up new possibilities for thought about the relationship between music and the individual How might these aims be realized? I propose two starting-points, from which creative practice can gradually diverge. First, following the examples provided by Sla˚ttebrekk (2010) and (even more innovatively) by Scott (2014), we take historical recordings from the very earliest years of the last century, when performance styles were less like our own than anything else we can hear, and begin to learn to play using the performance styles they document. That entails doing a lot of things that are more or less opposite to what is currently thought to be best practice. This is an excellent way of discovering how the opposite of orthodoxy can produce equally, even more powerful performances. Moreover, as one begins to see how not synchronizing parts, and not playing in time, and not sounding composed structure, not sticking exactly to the letter of the score, can all afford engaging expressivity, one begins to see how such techniques taken individually and pursued further, can allow one to find new, previously unimagined habits of performance style that work.
The second approach to freeing up our relationship with scores is in some ways opposite, anti-historical. This is to take scores and deliberately modify an aspect of received performance practice according to a systematic sequence of moves, in order to see what happens, to discover how each modification can (or occasionally, but rarely it seems, cannot) be made to work. An extreme form of this approach, but a very good one with which to begin, is that recommended above in connection with the Moonlight sonata; that is to say, take a score and do as far as possible the opposite of the expressive indications it contains. Fast becomes slow, loud becomes soft, and so on. Discovering how well pieces respond to this treatment is itself liberating. Proceeding more cautiously opens up a host of other techniques, including systematically placing emphases on different beats, different positions within a contour, or a text, or in a tonal hierarchy, and so on, testing one's ability at each point to make a counter-normative reading work; that is to say, to give it plausible musico-dramatic shape through expressive departures from a literal, machine reading of the notation (for that is what performance style, or 'musicality' does).

In conclusion
What we do with music, and what music does with us, becomes central to all of us once we understand the freedom -historical, ethical, technical -that performers have to make music mean. Performers must be allowed to share in the process of creating music from scores, in the satisfaction and the credit for creativity that accrues when they and their listeners recognize how much they have provided, and also in rights payments that are morally though not yet legally owed to them for their contribution which I suggest at least equals that of the composer. (Changing the law so that performers are no longer junior partners is a long-term project but one that it is time to begin.) If we could transform our attitude to performer agency along the lines I have suggested here, there just might be a chance that western classical music could offer more than comfort for the already comfortable, could be relevant to much more diverse audiences, could show us something challenging about ourselves, could be innovative, promoting not damaging performer health, encouraging not dismissing aspiring musicians creative enough to have ideas of their own.

Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The article is an additional output from the AHRC-funded Research Centre for Musical Performance as Creative Practice, award ref. RC/AH/ D502527/1.