PART 1: Performance style and what follows

1 Introduction and examples


Do an online search for performances of the Hallelujah chorus from Handel’s Messiah and listen to a few (try to avoid the ones you expect to like). There are performances with a semi-virtual choir of 2360; or with four soloists (accompanied by mandolins); performances by organ and brass; community choir, steel band and dancers; gospel choir; electronic organ and flash mob; and even some with small choirs and baroque orchestras that claim to be like Handel’s own. All of these, the full range (to judge by the comments added by listeners on sites like YouTube), inspire and move people intensely. And yet, in conservatoire and in the classical music business, only the tip of one end of this very broad spectrum is cultivated. For students of classical music, the only proper performance of this score is one that sounds just as Handel’s first performances are supposed to have sounded. According to this ideology, the job for which the classical musician is being prepared is to do history in sound. Anything else is wrong.

How musicians, critics and other ‘gatekeepers’—and knowledgeable listeners—respond to classical music performances depends to a large extent on beliefs about their propriety. If a performance is improper, in the sense that it is wrong as judged by history or tradition, then however lively, committed, engaging, fluent or technically brilliant the performance may be, the knowledgeable tend, at best, to disapprove, more often to dislike it strongly.

An aim of this book, however, is to show the danger of beliefs about what is ‘proper’ in classical music. While I do not address the huge field of current practices illustrated by this Handel example, I do want to broaden thinking around that tip of the spectrum currently occupied by respectable classical performance. I am going to suggest that there are many more possibilities there, and roundabout there, than we realise at the moment, and that musicians’ lives would be much more rewarding were they freer to explore them. At the same time, we may become more tolerant of practices that range more widely.


A more detailed example: ‘the Moonlight’

Under special circumstances, we do already allow a surprising diversity of approaches to certain kinds of scores. Mainly this applies where there are parallel traditions using modern and early instruments. Take the first movement of Beethoven’s ‘Moonlight’ sonata. In a typical current performance on a modern piano, the tempo is around 48 beats per minute, maybe a little slower. The music is quiet, calm, still. Alternatively, if you are an early keyboardist you play it on (a copy of) a fortepiano of Beethoven’s time. The timbres are very different: the fundamental is much weaker and the harmonics are much stronger and so the sound is brighter, more colourful, more dissonant in fact (because all those upper harmonics are very close together and clash in pitch). And the performance is usually somewhat faster, partly out of belief about Classical- and early Romantic-period speeds, partly because the sounds don’t last as long on a fortepiano and a sense of flow is easier to create if the gaps between them are shorter.

These two performances create different sound worlds, and to that extent different ‘Moonlight’ sonatas. In some ways they are more different than any two performances on a modern piano or any two on a fortepiano, although the latter are more varied than the former because to an extent fortepianists are still experimenting, both with the pianos, which are very different from one another (much more so than are modern pianos), and also with ways of getting these favourite, near-sacred scores to work with such a different mechanism and sound. (Compare Lortie and Beghin.) Yet both are currently acceptable in polite musical society, albeit with substantially different fanbases and for the most part practised by different musicians (though as time passes, and fortepiano-enthusiasm has come to seem less eccentric, there is increasing overlap). So you could say that there are two different performance traditions here, practised simultaneously but by different musicians with different beliefs about what is musically good. Each seeks to persuade their audience that they have reached the heart of this piece, that their Moonlight best evokes Beethoven’s. And indeed, whatever the differences among the performers, the same audience may very well be equally moved and interested by, and appreciative of both; though there, too, there will be those who think one more correct than the other.

As I say, this is a specialised situation; because, on the whole, most performances of classical music scores are as similar as almost any two Steinway performances of ‘the Moonlight’. Certainly there will be small differences, nuances of timing, rubato, loudness, from moment to moment; crucially also (on the piano) differences of touch. And these habits will vary from one player to another, the variation either celebrated or denounced depending on one’s target. (We shall look at critics’ responses to varying performances in chapter 9.) But nonetheless, no two Steinway performances will ever sound or feel as different as a typical Steinway and almost any fortepiano performance feel from each other.

And so the simultaneous practice of these different traditions seems to break the rules that maintain consistency in modern performance practice. That is to say, if you believe it is right that Steinway performances are as consistent as they are, then you should not believe that both traditions can legitimately exist at once. Or to put it the other way round, if you believe that both are producing legitimate performances of Beethoven’s sonata op. 27, no. 2, then you should also believe that performances can legitimately be this different within each tradition. But of course people do not believe that, and we shall need to examine why.


Older performances

Let’s now add a third and fourth performance, both around a century old, played by pianists born around 1860, who developed their way of being Beethovenian in the 1870s and 80s, nearer to Beethoven than to us. Arthur Friedheim (born 1859) was a pupil first of Anton Rubinstein in St Petersburg and then (from 1880) of Franz Liszt. His ‘Moonlight’ is very different. (Frederick Lamond (b. 1868), another Liszt pupil, plays it at a similar speed.) The tempo changes a lot, speeding up to settle in the general area of 72 bpm, though with much flexibility. Ignaz Paderewski (b. 1860) plays the score more slowly, typically around 54 bpm (a modern fortepiano speed) but again with much flexibility, and he does something else that would horrify a modern piano teacher but that was absolutely normal and considered sensitively musical by Paderewski’s contemporaries; he plays the left hand before the right, so each bass note anticipates the melody and inner notes written against it in the score. That one of the most famous pianists of all time should play in a way that could deny them a conservatoire place today shows just how relative ideas about musicality are. They change over time, and a thoughtful listener today is perfectly well able to appreciate them equally. Bear in mind, too, that scores composed contemporaneously with this kind of pianism (Brahms, Debussy, to name but two) ‘should’, if one actually believes in performing the composer’s intentions, be played like this today. So why is current practice so narrow? Why are historical models not followed in a tradition that purports to care about doing history in sound? And why is conformity to a much narrower agenda so rigidly policed?


A transgressive performance

Let us go further. It does not take a huge amount of imagination to conceive of a performance of this Beethoven movement that is as different from any of these norms as each is from the others. You could, for example, play this score much faster than anyone would consider proper at the moment. And that could produce a perfectly plausible musical result, where ‘plausible’ means simply that the sequence of harmonies and lines makes good musico-dramatic sense played with well-matched loudness envelopes, timbre and articulation. If you have a piano to hand, and play this piece, try it at 136 bpm, about twice as fast as the Liszt pupils, about three times the modern speed. (Or, to anticipate Part 3,  you can listen to Ji Liu playing it in Chapter 23.1.) Depending on the rigidity of your beliefs about what is proper, you may find that the notes are not made a nonsense of by being played ‘allegro’, or even ‘allegro’ and ‘fortissimo’: the piece still works as a sound sequence with melodic and harmonic coherence and dynamic shape. And yet it’s a reasonable bet than if you played the ‘Moonlight’ sonata like that in concert you would not be invited back. So what’s gone wrong? Why is everybody so cross with you for playing it allegro? What has happened that is not ‘musical’ or that is simply… wrong? An obvious answer would be that Beethoven wrote in his score ‘Adagio sostenuto’ and ‘delicatissimamente’. But given that he died in 1827, does he have a problem with our playing it allegro? Can he be hurt? (We shall look at the philosophical arguments surrounding faithfulness to the dead in chapter 11.)

We can get another angle on the normative thinking surrounding classical music performance by looking at a real example: real, in that a performer has dared to do it in a public concert (which at the time of writing is not quite the case for the ‘Moonlight’ as ‘Storm’, though I hope it soon will be).


Knowledge and belief

The violinist Patricia Kopatchinskaja has recorded Beethoven’s violin concerto with what on the face of it seems to be a very innovative first-movement cadenza. At this point in the score Beethoven has simply put a pause mark, and left it to the performer to improvise a cadenza of their own. Few (truth be told, probably none) take advantage of that today. There is a standard written-out cadenza by Fritz Kreisler and numerous alternatives by others. But Kopatchinskaja who, unlike her polite virtuoso colleagues, plays the violin as if her life depended on it, is soon joined by four cellos, and then by the timpani, then by the concert-master in a wild duet. It’s thrilling, and different in a performance of this concerto.[1]

Listeners’ reactions to this varied dramatically. For Andrew Macgregor, writing for the BBC’s music website, the cadenza ‘amplifies the sense of adventure and genuine re-discovery’, in a performance whose ‘soaring sound and improvisatory flair are compelling, and ultimately highly musical.’[2] (Note what different kinds of work that word ‘musical’ can do. We shall return to it in chapter 5.) Robert Braunmüller in the Munich Abendzeitung, sneering at her gender (‘eine Dame mit ihrer Geige’, a lady with her violin), is unashamedly hostile: ‘Patricia Kopatchinskaja deconstructs Beethoven’s violin concerto, as if it were Regietheater. Only, unfortunately, not as well.’ The cadenza was ‘a typical virtuoso insert including a duet with the concertmaster and four cellists’.[3] At the bottom of the scale, rondo1presto, a YouTube viewer, dismissed it as ‘a piece of shit’ and went on, ‘damn her with that childish cadenza’.[4]

Reading all the YouTube comments is, naturally, fairly depressing. The differences of opinion are as wide as it is possible to imagine, from Andrea Haubmann’s ‘Ich liebe Sie!!! Bin ein totaler Fan!!’ to JeffPuha’s ‘This woman should dig a hole and crawl into it to hide her shame.’ [5] But what’s very clear is that many of the enthusiasts are aware that Kopatchinskaja is drawing on Beethoven’s own cadenza written for his own arrangement of the score as a piano concerto. And many of the objectors are not. Similarly, Andrew Macgregor focuses his review on the alternative readings Kopatchinskaja has taken from Beethoven sources; Robert Braunmüller shows no sign of being aware of them, and I think we can assume that rondo1presto is not.[6] That the notes Kopatchinskaja is playing, while they may sound new, were actually written down by Beethoven seems to matter desperately to people: it changes their entire feeling about what just happened. On the one hand it was a fabulous musical experience, on the other it was shit. Yet the sounds in each case were exactly the same.

What does this tell us about the ability of people with musical knowledge to make musical judgements? The musically uneducated listener has no difficulty at all. The problem simply does not exist for them. For them the only question is, ‘was that a thrilling experience?’ If it was, it was a great performance. It is only the person with training who can hear a potentially thrilling performance and think it was shit because of what they believe to be a historical fact.

This is just as true of my hypothetical ‘Moonlight’ example. Once one knows that the score is marked ‘Adagio sostenuto’ and ‘delicatissimamente’, a performance ‘allegro furioso’, however thrilling it might be for an innocent listener, is shockingly wrong, a violation of deeply-held beliefs about faithful performance. The listener’s innocence, needless to say, is innocence of the composer’s instructions to the performer. And for most musicians, and most music philosophers and musicologists and music theorists and music lovers, the composer’s instructions are non-negotiable: they must be obeyed. This is a key belief, perhaps the key belief, on which so much else depends. And so the substantial differences in sound and performance style for the Steinway and fortepiano versions of the ‘Moonlight’ count for far less than the fact that both obey Beethoven’s instruction in the score.

By contrast, Kopatchinskaja’s performance of the concerto, for all its extreme dynamics, tone, rubato, and its unusual cadenzas, is in its broad sound qualities probably more like any other violinist’s at the moment than a fortepiano performance is like a Steinway version of ‘the Moonlight’, and we’re willing to accept both of those as legitimate. Yet hers is completely beyond the pale if we believe that she is making up the cadenza. (Jassim Khalil on YouTube comments: ‘You must be joking this is not Beethoven at all!!!’.) In each case we see that belief outweighs aesthetics. As long as beliefs are unchallenged, almost anything goes aesthetically: contrariwise, if beliefs are challenged, nothing does. (It’s no coincidence that scholars have from time to time remarked in print that they’d rather hear a bad performance on the right instrument, or using the right edition, than a good one that doesn’t.)[7]



We are beginning to see some fairly striking contradictions in these various responses to my examples. We can see that wide differences among performances are acceptable provided that all obey the composer’s score; but that very little, perhaps no difference is tolerated when the score is contradicted. (When we look at reviewers’ comments in chapter 9 we shall see just how unforgivingly that idea is enforced.) Yet clearly, if the score can validate both fortepiano and Steinway performances then there is something very wrong with using the score as a reference point, since the fortepiano performances and the Steinway performances we hear today cannot both sound any performance that Beethoven conceived (both may be very wide of the historical mark).

Yet it is not just the score that is regulatory for performance at the moment; it is also the ‘composer’s intentions’, of which the notes in the score form only a part. Together with the notes he left (about which there is often argument) go his conception of the structure (to the extent that that concept even existed in Beethoven’s time, which is open to question), and the way performers in 1801 touched the keyboard and moved from one note to another, that’s to say their performance style (which cannot be known until recording begins almost a century later). But in that case, if one can overlook the fact that none of these surely essential things can ever be known (essential if one believes that the composer’s intentions are regulatory), how could one ever allow a performance on a modern piano? And yet we not only allow them but find in them the some of the highest achievements of Beethoven performance (though not those, of course, with unsynchronised hands).

So clearly, when we want something to be acceptable we find a reason to allow it to be. In the case of the Steinway versions of Beethoven the usual idea is that Beethoven would have liked them if he had heard them. (Beethoven struggled against the limitations of his pianos and the Steinway would have been the answer to his prayers, etc.) But you cannot simultaneously believe that what happened in Vienna in 1801 is regulatory (i.e. you have to use a fortepiano) and that ‘what would have happened if’ is (or is even better). Or rather you can—people do—but it does not make sense to.



What is interesting, though, about the ‘he would have if only’ justification is its showing how essential it is to find an excuse in which the composer’s approval figures. But there is not and cannot be Beethoven’s approval for a modern, Steinway performance. It is simply convenient to allow it because it belongs within a practice that has developed up to this point and which we therefore pretend is traditional, justifying it (if pressed) through taste and through the pleasure given by the performances that are possible within it. In other words, by aesthetic, not historical criteria. And exactly the same goes for fortepiano performances. The piano may be contemporary with Beethoven, but the piano tells us much less than its players wish to believe about how it was played by Beethoven.

So what does Beethoven’s approval have to do with anything today? For reasons I shall set out in chapter 6, it is helpful for the maintenance of classical norms for young performers to feel that in some way Beethoven is present and looking approvingly on as they play. Without the composer’s approval it seems impossible for the classical music belief system—its ideology—to accept that a performance is legitimate, however exciting or convincing it might otherwise have seemed, or might currently seem to someone from outside the system whose response is only to the persuasiveness of its sounds. But where does Beethoven’s approval come from? How does he give it me? How do I know I have it? I know I get shredded by the critics if they think I don’t. But what happens to him? Is he himself harmed, now? How, exactly? The further we go, asking questions about these kinds of ideas and the contradictions they allow, the clearer it becomes that woven into classical music ideology is much wishful and muddled thinking and compromise and accepted contradiction and special pleading. And while a certain variety of thought is always welcome on any issue, the extent of these contradictions and (frankly) hypocrisies points to something profoundly disturbed about the ideology surrounding classical music.

The most evident consequence is that any kind of music-making that lies outside the narrow bounds of the ideology is, in effect, forbidden. If you play the first movement of the ‘Moonlight’ sonata ‘allegro furioso’, however well it may work with the notes, however fabulous your technique and your realisation of the idea, your reading of the score will be unacceptable within the culture. And so for Ernst Schliephake, what Kopatchinskaja does in the concerto (tame compared with playing ‘the Moonlight’ allegro) is simply ‘verboten’.

It takes a moment to realise the full implications of that idea. In what kind of society is an artist forbidden certain kinds of artistic creativity? Is policing necessary to generate the joy and fulfilment we experience from a great performance? Does it keep musicians happy? Or is classical music a kind of cultural North Korea? Later in this book there will be more to say about classical music in terms of this rather strong, alarming concept. For now, suffice it to note that Peter Cook and Dudley Moore skewered the policing of classical performance with characteristic precision in a sketch from as long ago as 1971 entitled ‘Prestissimo’.

DUDLEY [at the piano, playing Beethoven’s ‘Ode to Joy’ increasingly fast, stopped by Cook dressed as a traffic cop and holding a trumpet-cum-breathalyzer]: Ah, good afternoon, officer.

PETER: Good afternoon, sir. I’ve been listening to you for quite a while. You ignored two double bars, went straight through a coda sign and had a very nasty glissando when you were speeding up your fugato passage.

DUDLEY: Are you suggesting that I’ve been … playing incautiously? According to my metronome, I’ve only been going andante molto moderato.

PETER: Andante molto moderato? Very strange, sir. I had to go prestissimo possibile to keep up with you, sir.

DUDLEY: Well, I might have gone allegro con fuoco for a couple of bars, but I’m sure I didn’t get up to prestissimo.[8]

In the end, Dudley has to bribe Peter to let him off by buying fifty tickets and playing at the policemen’s ball. They are, of course, absolutely correct in identifying policing as an everyday response to illegal musical interpretation. Had Dudley explained that he was interpreting the score in his own way, Peter would no doubt have replied that it was not his place to interpret the law, only to obey it. Which is exactly what any and every performer has been told at some point by their teacher. Moore had studied at the Guildhall School of Music and Drama as a pianist, violinist and organist, and then read music at Oxford. He knew what he was lampooning.

What kind of artistic practice is this, then, in which performers spend around twenty early years of their lives learning to do just what they are told? Is it worth the effort at all, if that extraordinarily demanding training leaves one as little more than a mechanism for the performance of norms? And if those norms are riddled with contradictions, muddled and wishful thinking, as we have already begun to see, then who has one been forced to become? What values, whose values, do we perform as we ostensibly play our deepest selves through these remarkably potent scores? The erroneously imagined values of the imagined composer? Why should we faithfully obey anything that fabricated and confused?


Empowering performers

Much of this book is concerned with how to refuse to be policed, and about the benefits that could result for musicians and audiences if that kind of oppression were to be thrown off. It goes without saying that in many quarters this will not be popular. Many jobs in classical music are conceived and practised as faithfully ensuring that the imagined composer’s imagined wishes are performed as brilliantly and as persuasively as possible within the boundaries that define obedience. A lot of belief and self-belief is tied up in that process. Nothing threatens that as powerfully as a brilliant and persuasive performance of an alternative musical reading of a well-loved score, as Glenn Gould found with Bach, and Patricia Kopatchinskaja is finding at the time of writing with Tchaikovsky. The more powerful the alternative, the more hideously it threatens and the angrier the response.

This book aims to show the flaws in this kind of normative thinking, and to offer young professional musicians a way out of the straightjacket that norms attempt to impose, licensing much more varied performance in theory and offering models of how it can be achieved in practice. At the same time I shall argue that a more creative approach to playing canonical (and non-canonical) scores will bring benefits for musicians in well-being, prosperity and public esteem, and for audiences in fascination, revelation and pleasure. Most importantly, the book aims to empower performers and music lovers sufficiently for them to overcome the inevitable appalled hostility of gatekeepers to the profession (teachers, examiners, adjudicators, critics, managers, and the rest) who will have to face their demons at last, before sheer economic self-interest leads them to see classical music in a fresh light.

The consequences may well be hair-raising. Some of them were already set out quite a few years ago by Richard Taruskin in an essay that (perhaps not surprisingly) has been little cited thus far. His 2009 collection of essays, The Danger of Music, is perfect for dipping into, and that will have to do as my excuse for having reached its final essay only after several years of thinking and giving talks about the ideas set out now in this book. Taruskin argues there are no limits of principle that can be placed on the musical interpretation of a score, and that the worth of what a performer does with a score can only be judged by the listener (see also Chapters 6.19 and 19.2 here).[9] This may seem an insanely anarchic view to hold about classical music. But I hope as we work through the arguments that construct the case I make here, I may gradually persuade you that this is the only criterion that really counts. Or at any rate, if I fail at that, you may at least take away a more liberal view of what musicians are entitled to do when they use scores as a starting-point to make art with sound.


Continue to Chapter 2: ‘The fabulous status quo’

Back to Contents menu


[1] Other recordings (including those by Zehetmair, Kremer and Schneiderhan, and Tetzlaff) have used Beethoven’s piano cadenza but with less creative scoring.
[2] See also BBC Music magazine’s anonymous view: .
[3] ‘Patricia Kopatchinskaja dekonstuiert Beethovens Violinkonzert wie Regietheater. Nur leider nicht wie gutes. … Die Kadenz blieb als wildes Allegro eine typische Virtuosen-Einlage samt Duett mit der Konzertmeisterin und vier Cellisten.’
[4] Other YouTube comments invoking shit as the measure of this performance include MJWang and Arsen Stephanyan11, both at .
[5] Both at
[6] I did email him/her to try to find out, but I got no reply.
[7] See, for example, Malcolm Bilson, helpfully discussed in John Butt, Playing with History: The Historical Approach to Musical Performance (Cambridge University Press, 2002) 53–6.
[8] Willliam Cook (ed.). 2004. Goodbye Again: The Definitive Peter Cook and Dudley Moore. London: Century, p. 182. Later in the sketch: ‘Might I see your licence, please? … [Dudley hands Peter the score] Bit of trouble here, aren’t you? Ludwig van Beethoven: this expired in 1827.’
[9] Taruskin, ‘Setting Limits’, 453–4. Taruskin also predates points I shall make below about obligations to the dead and living, about opera production, and in asking who benefits from repressing performer innovation.

Leave a reply


  • Reply
    avatar image

    Greta Papa

    Sep 06, 2020

    Absolutely beautifully stated.

  • Reply
    avatar image

    William Drabkin

    Dec 28, 2020

    Reminds me of Schenker’s story about the proverbial composition tutor who tells his pupils, “When you are a Beethoven, you may write that way.” The safest way to escape the critical police is first to show that you can do something the right way, to become famous for doing so, and then to be imaginative. But perhaps that goes against human nature: we’re creative when we’re young and become set in our ways as we age. (Not you, Dan!)