This work is © The Reverend Gordon Giles 1995 (html version ©2001) and may not be reproduced in any form without the permission of the author. Any quotation should be properly acknowledged and referenced.

Before we begin, it is necessary to make some preliminary remarks, in order to delimit and explain the phenomenon which constitutes the broad subject of this thesis. Many readers will have a rough idea, gleaned from an interest in recorded music. The impression that such readers will have of the topic may well suffice, as I shall be leading the way into the world of authentic performance by reference to a paper in defence of authentic performance practice. But before I discuss any literature, I would like to characterise the phenomena which are commonly referred to as ‘authentic’ performances. I shall mention the claims of some of the adherents of ‘authenticity’, but it will be apparent that there is neither a complete theory, nor a great deal of consistency among the attitudes they express.

It is not my intention to ‘define’ authentic performance at this stage. Rather I want to ensure that the reader is aware of the subject under discussion. The notion of authentic performance today usually conjures up images of musicians playing in smaller orchestras than those to which we have become accustomed in this century, and playing on old instruments, or reproductions of instruments with which the composer would have been familiar. Often the sounds produced by such ‘authentic’ orchestras are recognisably different to those produced by ‘modern’ orchestras. ‘Authentic Performance’ is generally taken to be synonymous with ‘performance on old instruments’. More recently some conductors have persuaded ‘modern’ orchestras to play their instruments as though they were old but this has not generally been called ‘authentic’ - rather performances on modern instruments in older style have been called ‘authentistic’. This is presumably because we (or those who would guide us) wish to associate ‘authenticity’ with the use of old instruments, in which case a ‘modern’ orchestra can at best be influenced by ‘authentic’ ones - that it can be authentistic rather than authentic. In contemporary usage then, an authentic performance must at least employ old instruments (contemporaneous with the creation of the musical work).

Modern conductors who have associated themselves with this ideal include Christopher Hogwood, Roger Norrington, Nikolaus Harnoncourt, Terje Kvam, Roy Goodman, Paul McCreesh, John Eliot Gardiner, Sigiswald Kiujken, Gustav and Marie Leonhardt, Nicholas McGegan, Peter Philips, Andrew Parrott, Philip Pickett and Franz Brüggen among others. These conductors have been hired by major record companies, who have realised the great potential which the market for authentic performances has produced.

It would be easy however, to assume that authentic performance is this sense has been a recent phenomenon. This is only true as far as commercial authentic performance goes. The first ‘star’ in this field was probably David Munrow, who formed the Early Music Consort in 1967 and recorded discs for EMI, Deutsche Gramophon and Decca. He also had a successful television series on the BBC. To a great extent, the others have followed in his footsteps. But the idea of authentic performance as an alternative mode of performing music is not new. Musicians have not stopped playing shawms and sackbutts - rather the demand for, and notice taken of those who do has only increased in the last thirty years or so.

Thus it is not even true that it all began with Arnold Dolmetsch at Haslemere around the turn of the century. Wanda Landowska and Dolmetsch were not the first ‘authentic’ performers, even though they may have been the first to make this kind of performance practice acceptable and popular. Robert Donington seems to date the beginning of what may be called the ‘Authentic Performance Movement’ with "...the pioneering work of Arnold Dolmetsch a long generation ago." Joseph Kerman and Peter Le Huray use exactly the same turn of phrase: "Arnold Dolmetsch, the great turn-of-the-century pioneer..." and "His (Dolmetsch) The Interpretation of the Music of the XVII and XVIII Centuries (London, 1915) proved to be a pioneering work of enormous importance" (Dolmetsch was Donington’s teacher and mentor, and is the dedicatee of Donington’s seminal work).

But inasmuch as Donington, Le Huray and others date the invention of authentic performance to these early performers, we must acknowledge some kind of movement begun with Dolmetsch, taken to the concert platform by Landowska, and made universally acceptable by Munrow, Hogwood and Norrington. Generally speaking, then, it is the kind of activity undertaken by these people and their followers which is the subject around which I shall conduct my enquiry. ‘Authentic performance’ then, is perhaps best understood in terms of some kind of description of what these people do. Invariably, they produce public concerts and recordings of pieces of music - many of which are known to us already, on instruments with which, through their own efforts, we have now become familiar. The extent of use and popularity of such instrumentation has only recently become accepted and practised.

Introduction to the literature

In this section I shall briefly ‘map out the territory’ in which we find the term ‘authentic performance’ being used. The first point I wish to make is that the literature in the field, although it may seem copious, is generally not of a philosophical nature, and does not often demonstrate particularly clear thinking on the part of the musicians who write about what they do. In surveying this literature it becomes clear that much of it hardly represents a philosophical argument or defence of particular activities, and that most that is coherent is directed against the perpetrators of various forms of authentic performance rather than in favour of it. Much of the literature that deals directly with questions thrown up by authentic performers (by what they say and do), is of recent authorship, and appears in the Anglo-American musicological area. Writers of significance include Peter Kivy, Richard Taruskin, Stan Godlovitch and Charles Rosen. Other authors worth considering include the contributors to two anthologies, one which appeared in Early Music in 1984; the other edited by Nicholas Kenyon in 1988. Much of the more intelligent material has been produced either in these anthologies, or in response to them.

There are numerous books and articles devoted to historical performance practice - mostly aimed at musicians who desire to follow the old methods. Very rarely are the underlying assumptions questioned, except in separately collected publications and symposia. At a recent Royal Musical Association Conference, devoted entirely to authentic performance practice, no contributor discussed the relevance or validity of the activities which are conducted in the name of authenticity. The Journal Early Music is probably the most famous organ of such material, but the discussion generally centres around how things were (and therefore should be) done, but there is no demonstration of how we move from the ‘were’ to the ‘should be’. Material on performance practice is therefore not of substantial help to our enquiry about the nature and merits of authentic performance, interesting to musicians as it may be. There are whole books taking this kind of approach worth mentioning, such as the classic works by Arnold Dolmetsch, Thurston Dart and Robert Donington. Frederick Neumann has produced an unusual volume which offers some new perspectives on early performance practice, and an introductory section which is of interest to philosophers.

Some modern performers have written articles and books to defend their approach, but this material that is hardly philosophical in nature, and is often so confused that it is difficult to deduce what argument or justification is being made. There are books by Raymond Leppard and Nikolaus Harnoncourt which are almost exceptions, but it is still difficult to find a coherent explanation and defence of authentic practice. Most of the material in this category is to be found in popular magazines, such as Classic CD and Gramophone, or in newspaper articles. Such material is common, and is often fascinating, but it does not really aid philosophical enquiry.

Other writers have written overviews of the phenomenon of what they call ‘historical’ or authentic performance (often declining to see a distinction), setting out only to tell us what certain performers have been doing. Harry Haskill, Joseph Kerman and Norman Lebrecht have informed us where the history of historical performance is concerned.

There is, however, no single book, and very few articles that ‘do philosophy’ on this important musical issue. Lydia Goehr touches on the subject in the final chapter of her recent book, but even Michael Krausz’ anthology on the interpretation of music does not contain a paper on the subject. As far as I know, no-one has written a book on the philosophy of authentic performance, and papers on the subject are very few and far between. The appropriate literature is limited, but I shall turn to a discussion of it in due course.

Part One

 

Authentic Performance Defended

 

 

 

I have suggested that most of the literature on authentic performance is not of a philosophical nature; and that which is of most use is generally devoted to knocking down those who practice it, so to speak. Because of this, there is a danger that the latter group of writers become engaged in criticising their own construal of what is going on. Since no-one has really offered a secure argument or defence for the practice of using authentic instruments, those who wish to comment philosophically, have to first construe exactly what it is that they are dealing with. Thus they can sometimes make their task too easy, and criticise people unjustly; or talk in a manner which suggests that certain things are happening in the name of authenticity which are, in fact, not being advocated by anyone. As Charles Rosen says of Richard Taruskin’s views:

"Taruskin writes brilliantly and at the top of his voice, and his most crushing arguments are often reserved for opinions that no-one really holds. He asserts ‘To presume that the use of historical instruments guarantees a historical result is simply preposterous’. No doubt. Still, Taruskin beats his dead horses with infectious enthusiasm, and some of them have occasional twitches of life."

It is very easy to assume what the authenticists think, but there is a possibility that they don’t. Therefore, those who do the thinking sometimes have to create the positions they attack, themselves. I am going to try to avoid doing this, although I shall be turning to Taruskin’s and Rosen’s views later.

1.1 Stan Godlovitch

It is perhaps precisely because there has been so little written in defence of authentic performance that Stan Godlovitch approached the issue. We know who the authentic performers are, but they have hardly defended themselves. Godlovitch comes to their aid in two ways. He begins by objecting to some of the criticisms made of authentic performers (whom he calls ‘Purists’), by exposing weaknesses in these criticisms, and by attacking the alternatives which the critics establish instead. Then he goes on to defend the use of old instruments by making three main points. I wish to turn to these three points first.

1.1.1 Godlovitch and the ‘Purists’

These defensive points appeal to three concepts: intention, functional fit, and essential nature.

1) He characterises the composer’s intention as ‘what the composer wanted’. He construes what the composer wanted as expressed on the score, in the form of directions concerning instrumentation. The composer, in composing a work, has a specific instrumentation in mind. This can only be disputed in two ways, either by someone saying that the composer has not been specific (Bach did not specify instrumentation for The Art of Fugue, for example). Alternatively, someone may claim that a composer (Dowland in Godlovitch’s example) would have sanctioned the use of more modern instruments had he known about them. Such counterfactuals cannot be dealt with conceptually, and as Godlovitch points out, there is hardly any evidence for advocating such a hypothetical preference anyway. Presumably he would have difficulty with the Bach case either, because i) it is exceptional, and ii) that Bach did not express a preference (leaving aside the fact that he died before completing the work) suggests that he did not mind which instruments were employed and would have been happy for any combination known to him to have been employed.

Godlovitch thus maintains that the composer - as ‘maker’ of the work - has instrumentation in his mind from the work’s very conception - for him, "Composing is choosing... The centrality of the score is undeniable". Why should we not respect this aspect of the composer’s intention as manifest in the score, if we ‘respect religiously’ the pitch and rhythm he has indicated? For Godlovitch, the instrumentation and the notation are equally significant intention-manifesting constituents of the work as embodied in the printed score. We attribute privileges to the composer when we consider questions of notation, so we should also do so when we consider instrumentation.

The point is not so much that we should not alter the composer’s instructions in respect of notes and rhythm, but that we do not. We do not consistently do it with instrumentation either:

"Whereas we don’t seem to hesitate playing old harpsichord concerti on pianos, few would think it appropriate to play a modern harpsichord concerto like Poulenc’s on anything but a harpsichord. My suspicion is that there are no very good reasons for this seeming simple inconsistency. If it matters to and for Poulenc, it’s got to matter for Bach."

We can play on modern instruments of course, but the unusualness of Godlovitch’s position is that he wants to know why we should, whereas most critics of authentic performance want to know why he should want to play on an older instrument. He has turned the question around, and rather than ask why he should not do so (which is not quite such an interesting question), he challenges the assumption that playing on anything other than the instruments conceived of by the composer is being true to the work as set out in the score. It may be interesting to use modern instruments, but to do so is to issue a kind of insult to the composer’s intentions, because underlying our ‘modernisation’ is the assumption that the composer is not such a good craftsman, and that we are therefore ‘correcting’ mistakes.

When we choose to perform the work, we are dealing with the composer’s intentions, which we either respect, or vary. If we vary the instrumentation, then we must acknowledge that we are varying those manifest intentions. If we accept such variation, Godlovitch implies that other such variation must be accommodated, presumably concerning which notes we play, the speed we play at, or any other aspect which we take to be constitutive of the piece. He believes that authorship is significant, and that intentions are worthy of respect; and is obviously frightened of the proverbial slippery slope - for it seems that to tamper with the composer’s instrumental intentions is different only in degree to much more severe tampering: "Who adds figures to Rembrandt? Who alters metaphors in Donne?"

2) The composer’s instrumental intentions are not whimsical - they are guided by what is most appropriate. This appropriateness is determined by contemporaneous tradition. Thus the composer considers and selects from the resources available and tries to achieve a potentially optimum instrumentation, which he then specifies. The optimum, in this case, involves the assumption of most able performers and high-quality instruments. Whether he gets this is not relevant, if he does not, the optimum has not been achieved. Given the above point about counterfactual intention, Godlovitch must intend the optimum to be conceivable by the composer. Bach’s optimum, could not involve a piano. If one were to play Bach on a piano therefore, one would be departing from an attempt to produce an optimum result, which would itself entail a failure to execute the work properly.

Godlovitch says that it is possible to use the wrong instrumentation - and we should avoid doing so. Without stipulating exactly what is required, he makes the wise point that the ‘right mix’ is to be preferred, rather than underdoing it, or committing overkill. He believes that:

"Unless we have solid reasons to assume the contrary, the original scoring and interpretative directives are the closest we get to the right mix."

Hence Godlovitch believes that the instruments stipulated by the composer should be seen as the considered choice of the composer as being the most appropriate. Not only have we no good reason for ignoring the composer’s intentions about which instruments to use; but the composer happens to have selected the right ones anyway. This view comes close to one which suggests that the functional fit of the instruments is determined at least partly by the composer’s intentions, in that the composer intends a certain result to which he or she issues instructions on the score, and that the best arbiter of this ‘functional fit’ is the composer himself, because he or she is the best judge of how he or she intended the piece to be played. I am not so sure that this point of Godlovitch’s does not take is straight back to the point about intention, but I shall preserve his claims as he has presented them.

In defence of this point, Godlovitch pursues a point about originality, claiming that if pieces are conceived with particular sounds in mind, a variation in instrumentation must lead to results which differ from the original (what original?). By original, he may mean what he has earlier called the ‘optimum’. The point being, ultimately, that if one differs from the instrumentally-intentional instructions issued in the score, one will not have a chance of realising the composer’s desired optimum. If one does something which prevents such an optimum being attempted, then to a certain extent, one will have failed to execute the work properly.

This is the price to be paid if one alters or ‘updates’ the composer’s instrumentation. The original, according to Godlovitch, is lost, and the Purist values the experience at its best, and the best must be prescribed by its maker. He defends this view against objectors by referring to Plato’s opinion that it is best to leave cooking to cooks, and music to musicians. Plato would ask the composer how he wanted his music played, so it is better if we do so too. We should defer to the expertise of the craftsman (the composer), because the only way of not doing so is to behave in a manner which suggests that we think of ourselves as possessing superior expertise (or as having no respect for such expertise in the first place).

3) Godlovitch’s third point involves the concept of what a work ‘really is’. He believes that not only do note-sequence and rhythm constitute the piece as far as the composer is concerned, so too does its instrumentation. If the piece is the product of intentional activity, then its instrumentation has equal status as do the other features of its notation. For it seems that music is not just pitch and rhythm. This point hangs significantly on the first point about intention, but is supported by another claim.

The timbral features of a piece are part of its identity, and since different instruments produce different timbres (i.e. different series of overtones, leading to discernibly different sounds), the instrumentation makes a claim to be part of the essence of the work. If we use instruments other than those stipulated, we slide down another slippery slope - towards the pit of inappropriateness. We laugh at Chopin played on tubas, because it is inappropriate. For there are limits on appropriateness - some pieces have a wider "interpretative elasticity" than others. To play Chopin on the tubas is similar to playing Bach on a synthesiser or piano, the difference is only one of degree. Performances which are inappropriate in this sense are effectively pretending to be of the work, even if they are preferred by some people. To play Bach on the piano is to make an imitation, which can at best be reminiscent of Bach, like a kind of transcription. As Godlovitch puts it in conclusion:

"This isn’t to deny that we can come to like impostors and even prefer them over the real thing. But that doesn’t change their status."

So we are left with the three-pronged claim that 1) to vary the composer’s instructions concerning instrumentation cannot be justified; and that 2) to do so leaves us in a position from which we can hardly do justice to the work - 3) which is made up as it is of manifest intentions concerning the way in which a series of pitches should be rendered aurally.

1.1.2 The purists under attack

I think that there are some broader observations which may be made in response to Godlovitch’s defences of authentic performance. Ultimately, we cannot carry forward all his views into our debate, because I suspect that many of his points depend on each other.

In the first part of his paper, Godlovitch discusses some of the objections made against those he calls purists, and takes their side in attacking those who do not advocate authentic performance. Just as writing on the subject, this approach is refreshing, because very few people bother to defend themselves coherently against worthy criticism. As anyone would, Godlovitch chooses to defend his approach against certain arguments; and in doing so he issues some challenges of his own. He does not, unfortunately, set out opposing views to his own by reference to particular critics’ views, but I shall expound these criticisms as we move through the rest of Godlovitch’s unusual discussion.

At the outset, he sets out two necessary conditions for authentic performance:

"Any performance qua historical must be faithful to:

(1) the originally designated scoring and instrumentation where it exists, and

(2) the typical performance practices of the time (unless reason to the contrary is given; e.g. if a composer requires a non-conventional execution)."

He defends this stipulation on the grounds that historical information concerning (1) and (2) is no less easy to find than in other cases, such as the restoration of a Norman church. He thus seems to believe that questions about the availability of historical information, and the extent to which such information can be employed instructively are red herrings, which are certainly not any more relevant to authentic performance than they are to authentic anything else.

He lays out the debate in terms of three protagonists, Purists, Liberals and Hodiernists. The Purists are our authenticists - those who prefer historical factors, and want to implement them in modern performance practice as much as possible. Opposed to them are the other two groups, the Liberals want to permit many kinds of performance, both modern and authentic. The Hodiernists disagree with both parties, and want to defend modern ‘improved’ methods as being the best way to perform nowadays. Godlovitch then opens up the debate by taking the position of the Purists, who he believes, have been poorly represented, and whose ‘mission’ he seems to encourage.

1.1.3. The hodiernists

As Godlovitch admits himself, this term is rather ugly, but I suppose that ‘Modernists’ will not do. At any rate, he begins by making some of their objections, and defences of their own views.

a) Modern technological improvements

There is a view (which may now itself be historical!), that modern instruments are ‘better’ - technologically more advanced than those which they effectively replaced. Ralph Vaughan Williams was as good an advocate of this view as anyone:

"Opinions may differ as to the intrinsic beauty or otherwise of the harpsichord, but there can be no doubt that the pianoforte, with its infinite gradations of tone, from an almost orchestral fortissimo to an almost in audible pianissimo, performs the function of a continuo much better than the harpsichord, with its hard unweilding tone.

The same applies to our oboes with their lovely tone, which no-one hesitates to use instead of the coarse-sounding oboes of Bach’s time: why make an exception of the harpsichord as is now fashionable?..."

"It seems to me that to use the resources which we now possess reverently and with true musical insight is right; not only in the interests of the performers and hearers of our own time, but also as the highest tribute we can pay to the man many of us regard as the greatest composer the world has yet produced ... it is our privilege and our duty to use all the improved mechanisms invented by our instrument-makers to do full justice to this immortal work."

Vaughan Williams believed that if Bach had a modern piano at his disposal, he would have used it. The best way to play The St. John Passion involves modern oboes and a piano, because ultimately we should be interested in ‘doing justice’ to the work.

Godlovitch points up some weaknesses in this view. Firstly, it suggests that composers would "...customarily have wanted some instrument other than the one specified". That is, that they would always have wanted the latest ‘improved’ model, just as some people always want a new car or power drill. It also suggests that composers were frequently, if not typically frustrated by what they perceived to be limitations in what was available to them. There is no evidence for this, so it unreasonable to suppose that the composer’s intended specifications were offered in the hope that something better would come along one day.

Secondly, Godlovitch is unhappy with the view that we must always use the best possible resources (tools) for the ‘job’ of performance. He is worried about commercial influences which produce the most reliable, cheap instruments, which under this view, should then be preferred. This leads us to slip down the slope towards having to prefer electronic synthesisers, as being the ‘best’:

"So far from sanctifying the virtues of the piano over the harpsichord, the consistent Hodiernist must be prepared to throw the lot over in favour of the Compleat Synthesiser. Some may hail this eventuality, but it begins to do funny things with designations like ‘string quartet’."

Thirdly, it is not certain that modern instruments are necessarily ‘improvements’ upon inferior earlier ones. Some instruments have never been altered, while others have died out. A lute is still a lute, although trumpets have ‘evolved’. The modern trumpet is thought to be better by those who play it today; and the baroque trumpet is preferred by, well, by those who prefer to play the baroque trumpet. Thus, the claim to superiority may amount to a claim for familiarity. It is commonly the case that people prefer what they know. When people get to know about valveless trumpets, they may well end up preferring them.

Given the improvements to instruments (the adding of keys to wind instruments, or valves to brass instruments, for example), it does not follow; nor is it contingently true that the expressive qualities of these instruments has improved. The addition of keys, new designs and new materials, affects the timbre of the sound produced, which ultimately causes a performance to sound noticeably different. Godlovitch wants to know why these additions should be seen as necessarily producing better results, rather than merely different ones. Indeed, there is plenty of literature to suggest that the older, ‘unimproved’ instruments have more character of sound, and that a preference for them can be justified:

"The eighteenth-century bassoon, like the flute and the oboe of the same period, was more capricious and temperamental than its modern equivalent. But it was also more sprightly and plump, and the contrast between its different registers was greater than it is on the sophisticated, bland Heckel bassoons used in most orchestras today."

"Yet no-one who has heard Schubert’s songs accompanied on a good Viennese-style fortepiano, or Beethoven’s sonatas played on an early-nineteenth-century Broadwood, can deny that historical instruments bring out qualities in the music that are obscured or lost altogether in performances on a modern concert grand. In the hands of a skilful player, the fortepiano’s crisp articulation, evenness of registers and overtone-rich sound seem as ‘right’ for the music of Mozart and Schubert as the harpsichord’s brittle brilliance does for Bach and Couperin."

The point is, of course, that those who oppose authentic performers have not proved that modern instruments are necessarily better, and so do not have a strong basis for insisting upon their preferences. If they could prove that modern instruments were necessarily better, Godlovitch would hoist them on their own petard by presenting them with ‘superior’ sound producing synthesisers, which they would be obliged, philosophically, to play.

b) The inability to go back in time

Any performance we give, or attend, takes place in our own century, no matter how much we want it to ‘be’ in an earlier one. We cannot even perform yesterday, let alone two hundred years ago. Thus, it might be said, performance on old instruments is not ‘historical’ at all:

"...a reconstruction of the original sound is the most misleading translation because it pretends to be the original, while the significance of the old sounds have irrevocably changed."

Even if we produce the authentic ‘sound’ the effect on us of doing so would be very different:

"...as modern listeners can never hear with eighteenth-century ears, attempts at historical reconstruction (or so-called ‘authentic performance’) are preordained to failure. This argument maintains that, even if we could fully recreate eighteenth-century sounds, their effect on us would not be what it was on listeners two hundred years ago."

This attack is too broad, according to Godlovitch, who points out that it should not only be directed at authentic performers. Any attempt to reconstruct anything in the past is equally doomed to failure. We are who we are, and we cannot be anyone else. Why this fact should be specially wheeled out against authentic performers is not clear. That it is true, Godlovitch adds, means that there must also be an unassailable gulf between one person’s (sub) culture and everyone else’s, and this leads us down the path to solipsism. Having warned against this slippery slope, he then stands at the top of another:

"No-one can doubt, however, that when we use old instruments and former interpretive directives we’re closer to the spirit of the past than if we neglect these altogether."

We may not be able to measure our closeness to the past, but, even though he acknowledges that the criticism of historical nearness is as valid for authentic performance as for anything else, he wants to allow it to a degree, because it is his conviction that old instruments bring us closer to the past (even if we cannot actually get there), if only because they are old. If the extreme view is taken - that the past is unavailable - then Godlovitch wonders how the past can be understood at all.

c) The inherent inauthenticity of authentic performance

The attempt to be authentic, it may be said, renders the goal unachievable. The objection can be an interesting philosophical one, although Godlovitch does not deal with it in its most interesting manifestation. Performances which are based on detailed study of technical and historical details lack spontaneity, naturalness, and often seem self-conscious. The only way we can really be unself-conscious over performance is just to ‘be ourselves’ - so goes the objection to authentic performance - and the only way we can be true to ourselves is to do what comes naturally to us - not what came naturally to people of the past. The authenticity is really about unself-conscious genuineness, rather as Richard Taruskin has put it:

"Authenticity, on the other hand, is knowing what you mean and whence comes that knowledge. And more than that, even, authenticity is knowing what you are, and acting in accordance with that knowledge. It is what Rousseau called a ‘sentiment of being’ that is independent of the values, opinions and demands of others."

Not only is this authenticity in the best sense, for the Hodiernist, it happens to be the only kind which is attainable. Any other kind of performance will be stilted, according to Godlovitch’s construal of the criticism made of historical performance practice.

He has four answers to this point which the Purist might give. Firstly, he points out that an objection to stilted, or unnatural performances should not be (and is not) specially directed at authentic performers, the complaint is directed at stilted performances in general, of which some, admittedly are ‘authentic’ - but others are not. Stilted performances are not desirable however they are produced. Godlovitch thinks it is unfair to single out authentic performers in this respect.

Secondly, he does not like the idea - implied in the objection - that modern performances sort of happen without effort when a performer picks up a modern instrument. Everyone belongs to a performance tradition of some kind, and has to work at their performances very hard. The ‘Romantic’ image of the performer unconsciously at one with his instrument is not acceptable, but the particular objections to historical performance suggest such a naive construal to Godlovitch.

Thirdly, he wants to probe the notion of ‘natural’ and ‘unnatural’. Really, he says, we are talking about custom. It is yet to be shown that someone ‘reared intensively’ in particular forms of period performance cannot produce the same kind of ‘romantically’ unself-conscious result. What is ‘natural’ to a performer is the style and manners he was brought up in, so to speak, and there is no reason why someone should not be brought up in period practice such that it is natural to him or her. Furthermore, Godlovitch cites counter-examples - of performers who he believes have been able to avoid stilted performance - such as David Munrow. He was not just a lecturer telling us what it was like - he was in fact a true ‘performer’, even though his style, yet unstilted, is not of our time.

If by virtue of having taken place in the twentieth century, such performances by Munrow et al are twentieth century performances (and not authentic ones), then we should have to admit that aspects of sixteenth- and seventeenth-century style remain, and say that modern tradition somehow includes past traditions; in which case, there is nothing to revive, and no useful distinction to make, other than some kind of verbal one, which makes no practical difference. Hence what we called the old tradition is still contemporary, in which case the complaint about stilted performances being the rule rather than the exception is reduced to the same point as the first of these four - that the important point is that we do not value them - whether they pertain to be ‘authentic’ or not.

d) Changes in meaning of musical terms

Godlovitch puts the (rather obscure) objection succinctly:

"Suppose an older piece was a fast one - say it was marked ‘presto’ - but that fast playing then would strike us as relatively sluggish. To perform this piece properly now we would have to play it faster than it would have been played in the past. Otherwise it would not now come across as fast.... a chief condition now for an ‘authentic’ performance might be contrary to the very historical features of the piece we presume to recreate."

It might be said that authenticity does not reside only in what you do - it is also tied into the contemporaneous significance of what was done. But Godlovitch calls the objector’s bluff by demanding for evidence that tempo markings, etc. meant anything different in the past than they do now. As Roger Norrington defends his timing for the third movement of Beethoven’s Ninth Symphony:

""Like every ‘slow movement’ in Haydn, Mozart or Beethoven this one is not very slow... His description Adagio Molto has led some interpreters to take the movement much slower than 60. But it is safe to assume that the Adagio refers to the rhythmic (minim) pulse. Because 30 was a speed not available on Beethoven’s metronome he often doubled the unit for slow speeds... I think we should be extremely chary of claiming that Beethoven, or his metronome, must have been ‘wrong’. Everyone in the world, then as now, knew 60. It is the speed that large clocks tick at!"

The criticism Godlovitch seeks to address here suggests that playing the score as written might be authentic in terms of the rendered score; but - assuming that the experience for past listeners can be recreated or emulated in us today - to achieve this latter authenticity one might have to sacrifice the former, and actually go against historically authentic practice. Thus two potentially mutually exclusive construals of authenticity are established; but ultimately we wonder why they should be established in the first place, because the basis for establishing them is not proven. Where is the evidence that composers did not write what they meant? Godlovitch will not accept the view until some is provided.

Godlovitch thus provides four complaints that he thinks a ‘Hodiernist’ might make about historically authentic performances. The ‘Hodiernist’ prefers a contemporary approach, on the grounds that attempts at reconstructing older performances are likely to be stilted; the instructions of the score are not as reliable as we might think; modern instruments are able to produce a better sound; and that there is a paradox involved in the attempt to create an appropriate effect on today’s listeners. He counters these views by claiming that authentic performance should not be singled out for these criticisms; and that some of them are based on assumptions for which there is little evidence.

1.1.4 The liberals

More serious objections are provided by the people whom Godlovitch calls the liberals, to whom he turns his attention after fending off the less significant attacks of those who prefer modern practices. He sub-divides them into two camps - ‘permissible’ and ‘missionary’ eclectics.

1.1.4.1 Permissible eclecticism

This group does not want to insist that modern (hodiernistic?) performances are better, nor that authentic performances are. Their basic position is that both kinds of approach are equally acceptable. In doing so they effectively favour the status quo in as much as they do not favour any particular approach. It could be said that this a fairly conservative approach, or also that it is not as liberal as it seems, because such people would probably want everyone else to adopt their view. Godlovitch characterises them with the following arguments:

a) Transcription is consistent with past practice

The habit of playing pieces on instruments for which they were not originally intended is not only not a new phenomenon, it happened at the time of composition. Much music was transcribed at or soon after the time of composition, and of course, some music was not written for specified groups at all. Bach and Handel are renowned for re-setting their own music, and Bach had no qualms about pinching and re-setting Vivaldi’s. Thus to ‘transcribe’ a piece for performance on modern wind instruments (for instance) is to be consistent with early practice, and so is acceptable.

Godlovitch replies with the observation that we should not slip from ‘some’ transcription to ‘any’ transcription. Composers, as the original objection grants, were confined by the instruments available to them, so the spectrum of transcription was limited by the scope of the intentions which the composer could have had. We cannot conclude that a composer would have consented to a transcription on instruments of which he could not have conceived. This argument is very like Godlovitch’s objection to the claim that a composer would have wanted us to play on modern instruments had he known about them (posited by the Hodiernists, see above). The liberal’s view is slightly weaker - they only argue that a composer would consent to such transcription, rather than staunchly advocate it. Godlovitch’s response is basically the same: there is no empirical evidence for their condolence; and there is a metaphysical problem when we defend our claims with counterfactuals. Just as one could argue that a certain possibility might have been acceptable to the composer, one could also suppose that it had been consciously rejected.

Furthermore, transcription was not all that common. When it did occur, it was often done (often by someone other than the composer) in response to a particular demand. There are plenty of examples of this - Mozart rewrote his oboe concerto (K314) for flute merely by transposing in from C major to D major, and rewriting the solo part. According to his letters, he did not like the flute very much, and was responding to a commission. But this is not a case of a composer exercising some principle of liberality, rather it is a composer responding to the practical considerations raised by the need for a speedy response to an unattractive commission. It is mistaken, says Godlovitch, to attribute any greater significance to these rare examples.

b) Practical considerations

Some liberals point out that there are not a great number of people who can play old instruments decently. We should therefore not insist that a minority mode of performance should displace everyone else. It would be better to hear old music played on modern instruments than not at all! As Malcolm Bilson puts it:

"I have often heard it said by scholars and others interested in performance on early instruments that they would rather hear a great artist on the wrong instrument than a mediocre player on the right one."

Godlovitch relishes this position somewhat, in that rather than address the issue raised, he takes it as encouraging the Purists:

"The Purist would take this to be the best argument for redirecting musical training along historical lines to make up the deficiencies which this position acknowledges. If performance using modern resources is a necessary evil, it is no less an evil for that. And it can be undone."

He does not point out that today this view is plainly not true anyway. There are plenty of ‘authentic’ bands around (although it has to be said that there is a considerable overlap of staff among them), but the demand for authentic performers is being met, and the standard has improved considerably. In the early history of authentic performance, some of the noises produced on old instruments were quite appalling, but nowadays we have (re)learnt how to play them, to the extent that some people can no longer identify a performance on historical instruments in contrast to one on modern instruments. As John Eliot Gardiner puts it: "We’re no longer an eccentric coterie of mad vegetarians who can’t play in tune". Gone are the days when authentic performances are aurally compatible with a poor school orchestra, with scratchy violins and split notes all over the brass section.

c) Taste

We can listen to what we like. The argument goes like this:

"If, for us, the best way we have of appreciating any given piece is through the most familiar means, then that is sufficient and is warranted by those results."

In the name of taste, Godlovitch fears, anything goes, if someone finds it valuable. Along with the Purists he resists the view, mainly because there is no check on how far the permissiveness extends. It cannot be regulated, he seems to say, and would therefore not adopt this kind of view at all. Presumably he would want to regulate the exercise of taste to some degree. The position he is criticising does not do so.

It also reduces the piece of music to something merely to be appreciated, in whatever way suits its audience. Godlovitch wants to give some credit to the composer’s intention, as we have seen, so it is not surprising that he is opposed to a kind of view that tolerates anything. He is aware of the danger of satirising the position, such that ‘anything goes’ - and acknowledges that the liberals do not go so far, and would not sanction certain ‘exotic’ performances. But then, he wonders, how are the criteria for acceptability reached? He accuses the liberal of removing the significance of the composer’s intentions and the historical milieu of the composition, in order to encourage maximum appreciation. This, he claims, is an abandonment of the historical identity of a piece. Ultimately, he admits that the Purists want to say that the identity of a piece is affected by its setting, which must therefore not be removed. Godlovitch’s disagreement with this liberal view is thus not a refutation of it, he merely disagrees on a matter of principle, indicating two possible views concerning the essence of music, neither of which he discusses.

1.1.4.2 Missionary eclecticism

At this point in his discussion, Godlovitch wheels in the ‘Missionary Eclectic’ to defend the view that appreciation should be the goal of performance, hence that many kinds of interpretation are acceptable.

a) Reinterpretation for Greatness’ sake

Such a person might press the argument by saying that the sheer greatness of some music is reflected in its adaptability to different instrumentation (transcription), or different playing styles. Our respect for the composer should not be so limited as to neglect the potential for expression through the extension of interpretative possibilities. Great art is timeless, and is not tied to the performing conditions of its own century (one can imagine the hodiernist making a similar claim to justify the refusal of authentic performance). The concept of greatness should override questions of historical accuracy, as Michael Tanner puts it:

"...great music demands great performance... If you play a recording of one of the great conductors... then the sense of devotion to the work, passed as it must be through a powerful interpretative personality, is overwhelming."

At best, under this view, authentic performance can only serve greatness.

"What I am claiming is that no amount of score-studying and first-performance researching will help to make a performance adequate, let alone great, and the idea of the interpreter effacing himself in the interests of the work is paralysingly naive."

According to Godlovitch, here we have an attack on those who presume to set the standards of taste by their authentic activities. The missionary eclectic wants to increase the expressive and interpretative possibilities for a piece. But there is a danger of novelty for novelty’s sake here - another slippery slope, it seems - if one wants to play Haydn then one should respect his intentionally-governed instructions. If one does not, then one may be playing, but it has nothing to do with Haydn. Godlovitch does not like the idea of music being performed in a manner that has nothing to do with the composer. Extension of expressive possibilities is therefore not intrinsically desirable, and should not be pursued as an end in itself. The value of such extensions - new interpretations - is questionable in many cases.

b) Delimitation of interpretative possibilities

If we shy away from the ‘anything goes’ mentality we have to draw some limits. But these limits should not be drawn merely on historical grounds. There is a threat that the Purist’s approach produces some kind of authoritarian attitude which disallows creativity altogether. As far as art is concerned, the historicists’ constraints are arbitrary, and it is not healthy to allow any one group to dictate how things should be done. This view does not condemn authentic performance, but does deny its exclusive claim to be appropriate; and implies that to take authenticity too seriously runs counter to basic principles about what performance is for. Authenticity, if you like, is just part of the wider gamut of issues with which performance is concerned.

Good taste, say the liberals (according to Godlovitch), cannot be regulated or defined. An environment which promotes good taste does not exclude variety, nor attempt to organise creative endeavour. The desired environment is fertile, and not constricting. Authentic performance can thrive in such a climate; but so too can other approaches, even those which are utterly opposed to it.

Godlovitch does not respond to this criticism - rather he acknowledges it as the ‘greatest challenge’, and suggests that it is incurably romantic. He goes on to elaborate his defence of authentic performance per se, which I began this section with. For ultimately, he believes that the only response to the claim that authentic performance is no more significant than any other approach consists in making a strong defence for the authentic approach. Certainly, the onus has been upon him to produce one, but it would not follow that a good defence of authentic performance would force us to the conclusion that we should not play anything written before the twentieth century on our current instruments. At best, it seems to me, his defence could establish authentic practice as valid, but to establish it as universally superior strikes me as an incredibly arrogant and philosophical impossible task. It is to this that I shall turn in the next section.

1.1.5 Summary of Godlovitch’s views

But first, a recapitulation. Godlovitch, we have seen, makes responses to certain objections to the view that we should play on authentic instruments. Having divided the ‘opponents’ of authentic performance into three camps, he suggested that the first camp - the hodiernists, claim that technological advances and our inevitable distance from the past argue against authentic performance, which - at any rate - produces frequently stilted and unnatural performances. They also claim that the import of some gestures cannot be reproduced authentically, because they have actually changed. Godlovitch defends the advocate of authentic performance against these objections, with the main defence that authentic performance is not alone in experiencing these difficulties - which do not trouble us when we want to restore a church, or study Roman law. The notion of ‘improvement’ of performance is questionable where modern approaches are involved; there are plenty of ‘stilted’ performances which are not authentic; and it remains to be proved that past uses of notation were so different to what they are now.

He then divides two similar camps from one other - speaking of ‘missionary liberals’ and ‘permissible liberals’. He feels that the arguments of the former are somewhat stronger. Godlovitch dismisses the argument that re-arrangement of scores is acceptable because it frequently happened as being plainly not true, and those re-arrangements that did occur, did so for practical, rather than ideological reasons. The complaint that authentic performance is not a practical enterprise because we do not have the staff is taken, not as a criticism, but as a challenge to musicians to get and learn how to do it. It seems to me, that this has been happening, and that authentic performance is hardly a minority interest in the music world any more. The missionary eclectics then step in to say that we should accommodate all tastes, and include authentic performance as part of a spectrum of possibilities. Great art, the missionary eclectic says, can cope with this range of possibilities. It is better to expose artworks to potential bad interpretation, than to deny all but a small group the ability to interpret ‘properly’.

Godlovitch has no reply to this other than to present the arguments with which I began this discussion. He argues in defence of the composer’s intentions, which he believes are manifest in the score, just as the specifications for pitch, duration and volume are. This respect for intention also funds his claim that there is some kind of best ‘functional fit’ - that is, that some instruments are better for the job than others. His argument is not as clear as he may think, because this point is dependent upon the first - and both are rooted in the view that instrumentation is part of the composer’s specified intentions, and that such intentions ought to be respected. Godlovitch offers no defence for this assumption whatsoever. It is possible that his third point of defence is supposed to be one, but his concept of music is based upon the connection between the composer and the work:

"What identifies the piece for its maker is not only its note sequence but also its instrumental and technical setting given the conditions of its creation."

The crux of his argument then, is that the conditions of a work’s creation are of fundamental importance. Ultimately, he shies away from ontology, defending his position with the claim that if a composer wrote a piece for the lute, he intended it to be played on the lute, and we have no evidence for supposing he would have liked it played on a guitar. And because the composer knew what he was doing, and deliberately wrote it for the lute, it is very likely that the piece will suit the lute best, and for that reason also, it should be played on a lute.

This may appear obvious, but it is of little use philosophically, because it begs a very important question about the significance of the composer’s intentions; and to what extent we either have access to, or should listen to what he thought about the work and how it can or should be performed. A very important issue raised by Godlovitch, but not dealt with, then, is that of intention.

The other issue which he raises - again, without realising it perhaps, is the extent to which anyone can claim pole position for authentic performance (or for any brand of performance for that matter). The liberal position which prompts him to present his defence, makes no other claim than that authentic performance is merely a manner of performative interpretation, similar, but also distinct from other approaches. That Godlovitch feels that this point must be answered suggests that he believes that authentic performance is better than other approaches. Naturally, he feels he must justify this, although I have suggested that he has begged a major question in the process of his attempt. I do not think that his begging questions about the rôle of intention is the problem with his attempt to argue that authentic performance is always the best way to perform. Rather I suspect that he has failed to convince because such an argument cannot be presented. It is not possible to argue that one mode of performance is necessarily the best, because intuition, and philosophical precedence tell us that there are not, nor can there be aesthetic laws governing such things. This something to which I shall return in discussing the work of Frank Sibley.

1.1.6 A criticism of the authentic performers’ defence and a task

No-one has satisfactorily defended the claim that we should perform authentically rather than by any other method. Some people, notably Godlovitch, have implied a certain assumption that there is only one way to perform a work - and that it requires the use of authentic instruments. The liberal view with which he contrasts himself argues in favour of a multiplicity of interpretations.

There are four cumulative assumptions made by his position, each of which may be fruitfully considered.

1) Firstly, we have the assumption that the work is identified by or in its score.

2) A secondary assumption is that the notated score consists of notation-symbols for pitch, duration and volume and written instructions, such as Adagio or Con Sordini or Flute Solo or even just Clarinet in A. Based on this is the assumption that these verbal elements of a score are indicative of the composer’s intentions in respect of the performance of his or her composition.

3) Then there is the interpretative attitude that considers these intentions to be significant, and relevant to the way in which we perform the piece. If these assumptions are reasonable, then a defence of authentic performance such as Godlovitch’s may be adopted safely. Even if they are though, it remains to be shown that the Missionary Liberal is wrong.

4) For there is a fourth assumption - which is that if the others are acceptable (and they have been assumed to be), then an argument must be produced which shows that authentic practice is always to be favoured. This assumption amounts to the belief that there can ultimately be only one ‘true’, or proper way of performing a work. Some people might want to argue that such a result should be known as the ‘authentic’ way of performing piece X. In this latter sense of the word (which I have not used so far), ‘authentic’ is synonymous with correct. Hence there is scope for a terrible confusion, made worse by the fact that some authentic performers believe that authentic performance (in the descriptive sense) is truly authentic in some value-laden sense. This is, no doubt, why the practice became known as ‘authentic performance’ in the first place. The more recent tendency to use the phrase in the descriptive - rather than evaluative - sense may well amount to an acknowledgement that in order to make the claim that authentic performance is truly authentic one has to assume a great deal about many things, and that even if these assumptions are not individually untenable, to string them together to produce the intuitively desired result is rather complicated, and cannot be achieved with the kind of certitude that such authentic performers had originally claimed for the ‘correctness’ of their performances.

Since it has never been done, I intend to construct the kind of argument which these assumptions indicate. Each one of the four has its difficulties, as will be seen, and so I shall present both sides of each argument. It will be realised at the outset that due to the interlinked nature of the potential argument, a casualty on the way causes major problems. If, for example, we cannot show how the composer’s intentions ought to be taken into account, then it will not be particularly significant that the verbal notations of the score are equally part of it, even if the work is identified in the score.

1.2. The relationship between the piece of music and the score

I do not wish to get involved in a lengthy discussion about the nature of music. However, it is an issue which we cannot overlook, because those who practice authentic performance tend to assume certain things about the nature of music, and its relation to its creator. We have already seen that Godlovitch represents the view that the score is somehow a manifestation of the composer’s intentions concerning how he or she would like the piece to be played. Secondary to this is an assumption that the composer has a preference for the way in which a piece is played; and that that preference captures the composer’s preferred interpretation of the work. For we cannot avoid the distinction between how the composer would like the piece to be played; and how he or she would like it to be heard. The general assumption which we might make is that the composer has particular kinds of aural experiences in mind, and creates a score in order to communicate how best to achieve the sounds intended. It may also be hoped that a rendition of the score will bring about the creation of the sounds to which we are intended to listen.

This is a minefield of assumptions and distinctions, of course, so we would be best advised to separate them. I shall not deal with intentions in this section, since I shall be turning to them in the next section. In this section I want to deal with questions concerning the status of the score.

1.2.1 Nelson Goodman on notation

Today we use scores as source-documents for the allographic artform of music. We tend to take the score as authoritative where questions about the work are concerned. We have become used to referring to a score of a Beethoven symphony as the symphony. We believe, perhaps, that for any work of music there is, or could be, some kind of notated score. This score - relating as it does to a unique work - is also unique among the group of members of a notational system. Notation, according to Nelson Goodman, is what distinguishes allographic from autographic artforms:

"Let us speak of a work as autographic if and only if the distinction between original and forgery of it is significant; or better, if and only if even the most exact duplication of it does not thereby count as genuine."

Goodman contrasts autographic artforms with allographic ones - an allographic, or non-autographic work of art cannot be faked, for instance, as an autographic one can. Thus, music and literature are allographic; painting is autographic. The notational character of allographic artworks is significant, because Goodman wishes to count as genuine any instances which comply completely with a specified notated score or text. Because music involves performances spread out over time, the notational system is vital for the creation of instantiations of the particular piece, especially where more than one person is needed to make a performance. The identification of the piece is made through the compliance of the performance with the notated score. The score,

"... whether or not it is used as a guide for performance, has as a primary function the authoritative identification of a work from performance to performance.... every score has the logically prior office of identifying a work."

A feature which is not part of the notational system is not to be used in identifying the performance as an instantiation of the piece. A score, says Goodman, is a "class of copies or inscriptions which so define the work." For each work, then, there is a single score which correlates to a single class of performances. This link is unbreakable, and is held by the concept of compliance.

The notation itself is constructed from ‘atomic’ elements, which can be combined to form compound characters of varying complexity. A note indicating pitch, for instance, is a single, atomic character, whereas series of pitches, harmonies and rhythms are compounded atomic characters. Since there is no restriction on the number of characters which can be compounded, we should say that the whole score is a compound character in an notational system.

Through the compounding of atomic characters, the score can preserve the identity of the piece. It is able to do this because it conforms to five criteria for a notational system:

1) No inscription can belong to more than one character in the system. Goodman calls this ‘character-indifference’, saying that an inscription must belong to a character, and not to any other character. There can be no room for ambiguity. "no mark may belong to more than one character." Goodman calls this syntactic disjointness.

2) Marks which are deemed to be joint members of a character are copies of one another. It must be possible to distinguish between characters - characters which cannot be distinguished are copies of each other. If a mark does not belong to both K and K’ it must be possible to determine that it does not belong to each one of them. This requirement can be referred to as syntactic differentiation.

3) Ambiguity is not permitted. The compliance relationship must be invariant, any ambiguous inscription will hinder the decision as to whether a performance (for example) complies with it. An ambiguous character must be excluded, because, even if the inscriptions which include it are clear, "different inscriptions of it will have different compliants, some inscriptions that count as true copies of each other will have different compliance-classes." If either of these potentially ambiguous cases is permitted, then there is a risk that the identity-chain from performance to score; and from score to compliant-performance will not be preserved. Thus each character must have a unique determination.

4) Compliance classes must not intersect, for if they do, "some inscription will have two compliants such that one belongs to a compliance-class that the other does not." Thus, as well as syntactical disjointness, there must be semantic disjointness.

5) The final criterion stipulates that given K and K’ with different compliance-classes, and for any performance which does not comply with both, it must be possible to determine that the performance does not comply with K, or with K’. A compliant must be sufficiently different to enable us to determine whether or not it complies with a character. Thus, semantic differentiation is called for.

According to Goodman, then, these five requirements enable us to determine whether or not a work’s identity has been preserved among given scores and performances. The system works both ways - we can identify constituent qualities of a piece and its performances by reference to the score; and we can also produce a score on hearing a performance.

This is the kind of view of scores which I think Godlovitch has is mind, not necessarily because he has accepted Goodman’s point of view; but because Goodman seems to provide the kind of relationship between the score and the ‘work’ which many performers take for granted. It is easy to suppose that we should check off performances against the score, so to speak; and also that if we play the notes of the score, we are playing the piece. Both beliefs reside in the notion that scores are in some sense scores of the work, i.e. that through the score we have access to the piece. Such a link between scores and the piece is generally believed by performers to be something worth preserving, if only because their intuition tells them that the link is in some sense necessary.

1.2.2 A response to Goodman

I have suggested that this is an assumption on the part of performers - that works are identified in scores and that scores are identifiers of works. Goodman has argued that it is the case; we must discuss his view further, not least because an alternative view is a challenge to the assumption which forms part of the rationale for authentic performances. Firstly I shall raise some problems with Goodman’s view; and then we shall look at an alternative construal of the relationship between the piece of music, and the score.

Goodman says that compliance must be total:

"Complete compliance with the score is the only requirement for a genuine instance of a work.... full compliance with the specifications given is categorically required."

As Lydia Goehr points out, a performance at which someone sneezes, must then be taken to have violated the "condition of perfect compliance, rendering the event something other than a performance of the work". She also explains that, according to Goodman’s theory, that two performances can and do sound different, because there are contingent features, which do not affect the compliance conditions; but this itself implies that:

"...any feature can be added to the sound event if it does not get in the way of the constitutive features being compared with; the consequent event is still a performance."

She wonders if this is an undesirable conclusion. Goodman would probably say that compliance is more significant for compound characters in the notational system, than for individual ones. A compliance with each individual produces compliance with the compound, and hence the sneeze is not taken to spoil compliance with the compound but with one atomic character. More reasonable perhaps, would be the idea that the sneeze is one of the contingent features which affects the quality but not the identity of a performance:

"Where violinists choose to add a little extra vibrato, so they might choose also to add a sneeze."

To counter this, she suggests a distinction between notationally contingent and accidental features. This distinction accounts for the element of choice which she has introduced - accidents are reckoned to be disruptive, and unintended by those performing - hence are irrelevant. The performer exercises judgement where notationally contingent features come into play.

But ultimately, she thinks that Goodman does not need these distinctions to help his case. Accidents do happen, of course, and we should take them as practical intrusions, and therefore judge performances which include them (albeit unintentionally) to be imperfect or incorrect. There is then, a gulf between theory and practice - sometimes a performance which is practically imperfect needs to be written off, but its being so does not affect the issue of whether perfect compliance determines the required identity for the performance in question. Thus an intrusion such as a sneeze is not a problem for Goodman’s theory, so long as we do not allow the practical considerations concerning adequate performance to interfere.

But it may be that we want to continue to consider performances with one or two ‘mistakes’ to be performances of a piece. Goodman manages to accommodate extraneous sounds like sneezes which do not upset compliance conditions; but he will not consider a performance with a wrong note to count as a performance which instantiates the piece of music:

"Nevertheless, the composer or musician is likely to protest indignantly at refusal to accept a performance with a few wrong notes as an instance of a work; and he surely has ordinary usage on his side. But ordinary usage here points the way to disaster for theory."

Goodman’s defence of this position rests in fear of the slippery slope by which all performances might be of the same work. It would not be possible, he suspects, to draw the line as to how many mistakes would be permitted before we say that the performance is of something other than the work of which it purports to be a performance.

Goehr concludes her discussion of Goodman’s theory with the observation - which I think is correct - that he has not considered the fact that performers intend to perform certain pieces:

"If performances of a given work do not end up exhibiting exactly the same (constitutive) properties, they none the less are identified as being performances of the same work in virtue, perhaps, of the work which they intend to instantiate and which we recognise them as instantiating."

But Goodman cannot accommodate this kind of view in his own, and retain the idea that scores and pieces of music are completely determinate where their constitutive properties are concerned. As we have just seen, he is ultimately content to ignore our intuitive understanding of musical practice. But then, we end up by having to conclude that the gulf between theory and practice is too wide to be glossed over.

Goehr goes on to merge the element of intention with Goodman’s views, in order to satisfy this major flaw. We can, she says, include an intentional element without making a circular argument. Someone with sufficient skill and knowledge could intend to perform a piece, in which case the condition that a performance is of a particular piece if and only if it is intended to be so, is satisfied. But while the appropriate intention is necessary, it is not sufficient. It is possible, she says, to fail to bring about a performance, even if one has the requisite abilities. So we need to add the condition that the performance be recognised as a performance of the named piece:

"Thus, something is a performance of a work if and only if (a) it was intended to be such a performance and (b) it is recognizable as an event intended to be a performance of this work."

Although this view allows a few wrong notes in a performance, it also allows for every note to have been wrong! To counter this, Goehr introduces a modified claim that there should be some compliance between work and score. A certain vagueness must be maintained, for we now demand that a performance be intended and recognised to be such, but must also be "fairly close to complying perfectly with the relevant score." But now we face the problem that the Goodmanian view sought to avoid - where do we draw the line? Goehr does not proceed to answer this question, for it is her major thesis that analysis ultimately lets us down when we ask questions which seek to define in this way. She believes that:

"Analysts who have sought to describe musical works have employed methodological principles and assumptions that impose unnecessarily severe limitations on their theories.... For this design has created an irresolvable conflict between theory and practice."

Goehr’s offer - that a performance be of a work if it approximates perfect compliance with the notational score and be intended to be recognised as doing so may be a disaster for theory; but it is neither unworkable nor non-sensical. While it cannot be defended analytically, it may be widely held. I suspect that hers is the kind of view that is held uncritically by many performers, some of whom carry the position forward when they attempt to justify authentic performance. It is not my purpose here to defend her view, or to continue the debate about the relationship between work and score. Rather I am concerned to indicate how a view such as the one with which she finishes the first part of her book underlies much thinking about how performances should be, and are, undertaken. Goodman’s criticism that theory is a disaster for practice; and Goehr’s that philosophical analysis cannot provide the solution, do not interest the proponents of authentic performance. It would be hard to see how they could, because Goodman admits that practice is at odds with his claims; and Goehr’s view argues against the adoption of philosophy to defend one’s position in the first place. If analytical philosophy cannot do the business, then presumably there is no point in trying to make it do so. Thus we cannot complain if the performers continue to hold the assumptions which they do - at least, they might say - what they do works in practice.

1.2.3 An argument against the assumption - Roman Ingarden

If one cannot resolve the problem of the work and its identity analytically, we can hardly criticise performers for the tacit assumption that scores are to be identified with the piece, and vice-versa. As we have seen, Goodman’s view is not far removed from this. There is not a strong view to defend this assumption, but we should now investigate the grounds for saying that we should not make such an identification of the work with the score.

The question concerning the identity of the musical score has also been addressed by Roman Ingarden. For him the score symbolises the music in virtue of its being intended to do so by the inscriber. The marks on the page are ‘designating functions’, which:

"...are intentional, imposed upon the sign by a particular operation of consciousness performed by a certain conscious subject. No physical object (a chalk mark, a desk, a lamp, etc.) can either designate on its own account, or effectively acquire a designation through the act of a certain subject. A physical object (a lamp, a desk, a drawing) can only be an ontic base of a sign, which itself is the product of a subjective conscious operation that grants it just such an intentional function..."

This is not to say that the signs on the page are of fixed size, or even of fixed shape. For as long as ‘conventionally agreed-upon aspects’ remain observable, the sign remains the same and performs the same intentionally designated function. A given musical sign, such as a note (k), has various features, but there is what Ingarden calls a ‘typical aspect’ which performs the intentional function. Thus, in a musical score, k and k perform the same function - that of being the sign for a note of a relatively defined duration:

"The so-called letters of the alphabet may be physically larger or smaller, a railway signal more or less covered in grime, and so on, but as long as the conventionally agreed-upon aspects may still appear palpably on the basis of real objects despite these variations, then the sign remains the same and performs the same function of designating."

The score now appears to be a means to an end. But it is commonly objected against the claim that scores are necessary for performance, that some performers do not use them. Two cases spring to mind - firstly, where performers somehow know what to ‘do’ without looking at a score while they do it; and secondly, where the music is composed and performed at the same time. For both cases a score is possible, in that after any musical event, the notes may be written down in one of various notations. That a performer does not use a score does not affect the fact that anything he plays can be written down according to a notation in which ‘conventionally agreed-upon aspects’ are observable. Memorising a score does not render it useless or unnecessary - indeed, if it has been memorised, the score has been the means to that end. If anyone wishes to perform the piece, he or she needs either a very accurate memory, or a score. It is possible to memorise a piece without memorising the score, which suggests that it is not scores which are memorised - rather it is that to which the score has an intentional relationship. This points to a distinction between score and music.

We can construe the score as little more than a set of instructions, and to admit that the same information concerning what should be done in performance may be conveyed by other means. Not only are there various notations which may be utilised in a score, the information may be conveyed without any form of inscription. I am thinking here of the way in which someone may hear a performance, and then be able to perform it afterwards. This is a rare ability, but is one said to have been possessed by Mozart, who wrote down Allegri’s Miserere after hearing it in the Sistine Chapel in 1770.

Ingarden says that we ‘fix’ the meaning of the notational symbols we use by attributing ‘musical’ qualities to the marks on paper, and we must be aware of the convention by which we do this in order to read off such qualities. By this method individual symbols have come to determine the pitch and duration of particular sounds. The ordering on paper of these signs or symbols indicates the way in which such sounds should be associated, and so the symbols not only tell us what the music should be like, they have an imperative function, in that they are instructive:

"But the role of the score does not end here. It also constitutes an arrangement of instructions as to how to proceed in order to achieve a faithful performance of a given work. In this double sense a score consists of ‘imperative’ symbols. The score, finally, is a way of revealing the composer’s wishes as to what the work is to be like."

Ingarden's account is grounded in convention - the musical signs which he discusses have a function in respect of a convention to make one thing (a sign) indicate another thing (a particular sound). A correlation is created, but is contingent, in that it would be possible to inscribe a piece of music in more than one way. For example, folk guitar players use a form of notation which is different to the notations used by classical guitarists, yet the same sounds can be indicated using either system of signs. Both systems have arisen conventionally, but the signs themselves are different.

Not only can pieces be notated in different ways - they can have features which are not features of the score, even if we allow for the intentional function of the features of the score. Ingarden claims that tonal and melodic qualities are not qualities of the score. These are qualities of the music, though, so it would appear that there are qualities of the music which are not qualities of the score; and there are also qualities of the score which are not qualities of the music. Thus Ingarden claims that the score and the musical work are separate, although the score is an intentionally designated object in virtue of its existence as a result of the intentional activity of musical composition. The score is not necessary though, since intentionally creative activity, such as composition, can result directly in a performance.

Thus, for Ingarden, in contrast to Goodman, the score and the music are to be kept separate. The score is intentionally designated with symbols which determine pitches; but those pitches are themselves not part of the score. The score is a set of instructions, conveying to the performers what they should do in order to realise the composer’s intentions concerning what the piece should be (sound) like. The score is not to be identified with the work, even though a performance of the work might be very difficult (but not impossible to produce) without one. For Goodman, the music is identified with the score - for Ingarden it is to be distinguished from it.

1.2.4 Comments on Ingarden

It is trivially true that there are different copies of scores, and that each copy has certain features which it does not share with all other copies (such as the kind of paper used for its production, and so on). These features are not features of the music, and the music and a given score can be distinguished easily. But this is surely not the point, because we are still left with the possibility of some kind of ideal score, or type of which the copy on the shelf is a token. Pieces of paper do not have melodic qualities, but this is a trivial point to make. Ingarden is unclear as to whether he considers features of the score to include features of a specific instance of a score (such as the paper it is printed on), or to be those features which would be found on any edition of a score. Since he is claiming that a score is made up of intentionally significant signs, we might assume that he means the latter. There is a possibility that in talking of ‘the score’ he has confused a given instance of a score with the concept of a score. ‘The score’ could be taken to mean either.

The notion of an ‘ideal’ score can be elucidated through reference to the type-token distinction:

"Those physical objects which (as we have seen) can out of desperation be thought to be works of art in cases where there are no physical objects that can plausibly be thought of in this way, are tokens. In other words, Ulysses and Der Rosenkavalier are types, my copy of Ulysses and tonight’s performance of Der Rosenkavalier are tokens of those types."

The score of Verklärte Nacht, is a type, in that it has tokens (such as the copy on my shelf), which need not share each and every quality. There are various editions of the score, and different copies of various editions have different features, such as having a library stamp on page 19, or a coffee stain on the cover. If we wish to consider what features the tokens of the type (score ) have which make them tokens of that type, then we can think of Ingarden's marks which are intended to signify certain sounds and performative actions. Library stamps or coffee stains have no such function, and so they are not features of the score, though they can be features of particular scores. These features are not essential to either particular copies of a score, or to a conception of the score itself. Ingarden's claim appears to be that because coffee stains are not intentionally significant functional features of a score, an individual score cannot be identified with the music, because such a feature is not shared by both.

If our notion of the score involves an ‘ideal’ score of which we have physical instances, then we can admit that these physical instances, or tokens of the type which is the score, have qualities which the type does not share. The coffee stain and library stamp are such features. There may also be features which the type may be said to have, but which a given token is said to lack:

"First we must obviously exclude from consideration properties that can pertain only to tokens (e.g. properties of location in space and time), and equally those which pertain only to types (e.g. ‘was invented by’). When we have done this, the situation looks roughly as follows: Classes can share properties with their members (e.g. the class of big things is big), but this is very rare: moreover, where it occurs it will be a purely contingent or fortuitous affair, i.e. there will be no transmitted properties. In the case of both universals and types there will be shared properties. Red things may be said to be exhilarating, and so also redness. Every red flag is rectangular, and so is the red flag itself."

Talk of types and tokens can be confusing. For that which we call the type of which individual scores are tokens, need not be called a score at all. Generally we mean to refer to something physical when we talk about scores, and so to talk of a non-existential score is contrary to common parlance. We might as well call such a non-existential score ‘the music’, since we find that if we talk of ideal and real scores, as I have been doing, the question as to what constitutes music is transmuted into the question as to what constitutes scores. We can claim that a coffee stain is not essential to either a given, or notional score, and we soon find ourselves saying that the coffee stain is not essential to the music either. It is, however, an feature of a given score, but it is not a feature which has anything to do with qualities of a score which bear some relation to what is heard or written down.

The score need not exist in a physical form. This is to say that there need not be any physical object which we call ‘the score’. Ingarden points out that there are plenty of pieces of music which have no score; and we would not want to say that if every copy of the score of a certain work disappeared that it ceased to exist as a piece of music. We do not invest a single object with the status of score, one of many copies (or even no copy) will do for the purposes of performance.

Goodman’s view on the relationship between scores and pieces was counter-intuitive, by his own admission. Goehr added the (non-analysable) stipulation that the performer should intend to be performing that work, and that he or she should be recognisably doing so. She modified his claim, suggesting that perfect compliance should be more or less achieved. I have suggested that Goehr’s modified view covers the kind of attitude which modern musicians have towards performance. I also suspect that Ingarden has something to contribute on this matter. His view that the score constitutes a set of instructions from composer to performer must be right - at least in that many performers treat scores as having that function:

"First of all, we try to play the score (of course with all its repeats). Since it is immensely detailed and the work of a genius, it does not seem necessary to add all sorts of extra speed changes or to alter those that Berlioz prescribes... If all these features make perfect sense of the emotional world of the Symphony, so too do the innumerable exact details of note lengths, phrasing and dynamics... We find that the more accurately we reflect such gradations, the better the music sounds."

We have seen that Goodman does not believe that the score’s function as a guide to performance is of primary importance. Both he and Goehr gloss over this function. This is a pity I think, because not only do performers tend to treat scores as serving and instructional function, composers do produce scores at least partly for this reason. It is true that scores are not necessary for performance, in the theoretical sense, but when a group of musicians get together they have to resolve certain practical problems. Access to a set of instructions is beneficial; and the score (or parts copied from one) is the most practical solution. Scores are not defined by their use, but it does seem remiss to ignore this use, and, as Ingarden points out, scores are intended to have this function. Goodman disagrees with him over the primacy of this function.

1.3 The score and verbal instructions

The second of our original assumptions is that the verbal instructions, issued by the composer are part of the score (which identifies the work). Goodman believes that tempo markings, and other such ‘verbal’ instructions, are not to be identified as constituent parts of the piece of music, because they not notational. Consequently, where performances differ in these respects (as they invariably do), no problem is presented, because these are not features for which compliance is required.

"Thus the verbal language of tempos is not notational. The tempo words cannot be integral parts of a score insofar as the score serves the function of identifying a work from performance to performance."

The quality of the performance of a work may be affected by different adhesions to specifications, but since these specifications are irrelevant to the identity of the work; they cannot be used to challenge a description of that identity. Goodman is not interested in how performances come about, or how we read scores. He is concerned mainly to specify conditions for scores, and for performances. In both cases, the relation between the notation and its compliants must be complete:

"... a performance, whatever its interpretative fidelity and independent merit, has or has not all the constitutive properties of a given work, and is or is not strictly a performance of that work, according as it does or does not pass this test."

Ingarden, arguing from a different corner, would not want to distinguish between the intentional functions of the notes, and the verbal instructions. Both enable communication to take place between composer and performer. Notation, it may be said, enables the composer to specify which pitches should be played in which relations; while verbal instructions serve to indicate tempo, mood and volume (ff and p etc., are not notations, they are abbreviations). As such, both are intentional, communicating an intention in the form - ‘I, the composer, want you to do X’. Here X is an action which is conventionally associated with the performance of music. When the composer writes an F-sharp on a score, or in some other way indicates one (through verbal, spoken instructions perhaps) he or she is specifying a pitch which he or she intends to be performed in the context of the other pitches. This context is also established by the composer through the specification of other pitches and instructions concerning how he or she intends them to be related to one another in an aural environment.

It is clear that questions about intention are being begged. I shall be turning to these questions in the following section.

The point here is that Ingarden, who does not hold the view that qualities which are not notational are not of the music, is not forced to agree with Goodman. Following from Ingarden we do not have a problem in saying that notated characters and verbal instructions have the same communicative function of instruction. Goodman is tied to this because he believes that the score’s primary rôle is to identify the work. If we can get anywhere with the kind of alternative that Goehr proposes, then we are not tied to the belief that the instructional aspect of scores is not relevant.

Goehr suggests that if character-compliance is considered not in terms of conditions being met but in terms of ideals being striven for, then we need not be confounded by questions of delimitation. The Goodmanian view proposes that compliance is a necessary condition for a performance - Goehr resists the idea that compliance is a necessary feature - she prefers to think of it as an ideal. She believes that the incompatibility between the theoretical stipulations associated with the identity conditions; and the phenomena of performances, demands that the Goodman view be rejected or modified. In her book she proposes that the ideal of character-compliance is a recent ideal, and only characterises Western Classical music:

"What we understand today to be perfect compliance has not always been an ideal, and might not be in the future. Actually it is quite peculiar and rather unique. It has characterized classical music practice only for the last 200 years. It is also not universal in the world of music... Whereas in classical music performances we strive towards maximal compliance with a fully specifying score, in traditional jazz improvisation, where very different notions of compliance operate, musicians seek the limits of minimal compliance to tunes or themes."

While working with ideals removes some of the problems brought along by the quest for necessary conditions, it does introduce a tension between what is sought and what is possible. But all this tells us is that we cannot insist that performances be compliant with the score - rather we should strive towards such compliance, in full awareness that the ideal cannot be attained. The recognition that something is an ideal does not detract from its existence as such.

If perfect compliance is an ideal, rather than a necessary condition, then it cannot be demanded - it cannot be a prerequisite. Furthermore, says Goehr, the notion of an ideal enables us to individuate performances. We no longer need to talk of a strict notation which identifies the work in a chain of score-copies and performances; we can propose instead that the score be as accurate as possible and that the performance be an attempt to live up to an ideal (which is never achieved).

Our conception of the score as instructional can be brought in here, because once we have an ideal, we have something which is action-guiding. The ideal serves as an arbiter of normalcy -

"...we are guided by certain beliefs and we develop appropriate skills. All of this comes to be reflected in the institutionally generated expectations that are bound up, for example in the musical world, without producing and recognizing events as performances of musical works."

If Goehr is right, then scores are created, offered to the world, and received as being a set of ideal instructions. Perfect compliance is not possible - which we could take to mean that a ‘genuine’ performance could never be given. All performances must fall short of the ideal; but, this being the case, we do not make such a stipulation for performances, rather we suggest that all attempts to realise the ideal are performances. Goehr would probably remind us to add that such attempts are performances for as long as they are recognised to be attempts at realising the perfect compliance conditions for the piece in question.

Since we have moved away from Goodman’s conception of scores and performances, to one that speaks of ideals, we need not be tied to his claim that performances either comply with the score or with the supplementary verbal markings. It seems to be possible that they could comply with both. We cannot allow that they only comply with the verbal instructions, because such instructions do not concern which pitches should be played. Rather, they give supplementary guidance, concerning how those pitches are to be rendered.

It might be said, though, that instructions concerning how a pitch is to be rendered can affect its frequency, even if we cannot perceive that change, except perhaps when two pitches are compared with each other. Thus we can say that the way in which a pitch is indicated, perhaps with an accent-mark, or tenuto, or sffz indicates a frequency, the perceptual identification of which we cannot actually make. That is, that we cannot determine whether compliance conditions for such a supplementary mark are met. We can say that the compliance conditions for it being an A natural are made, because we perceive it to be A natural. But ‘A natural’ may cover many frequencies between 400-444hz, which, independently of a comparative test, we are not able to distinguish. We take different frequencies to be compliants of pitches, when we cannot actually tell if they are such. As a result we may be in a position in which we cannot actually tell whether a sound event complies with the supplementary instructions or not. If this were ever true, then the conditions for the determination of compliance could not be met under normal circumstances; and we would not be able to determine whether a performance was compliant or not.

According to Diana Raffman, Goodman has made a crucial mistake in forcing the distinction between notational markings and verbal instructions. It is only trivially true that a performance complies with the verbal instructions, because it also complies with the notational markings. Goodman must mean that the performance as a compliant instance of the piece is distinguishable from the performance as sound event; the latter of which is determined by compliance to sound-event classes (rather than work-instances). But Raffman accuses Goodman of having confused sound events with acoustic events:

"In identifying the compliant of the supplementary markings with the performance taken as a sound event, Goodman illicitly transports features of the acoustic realm into the perceptual; specifically he mistakes a dense ordering of frequency-classes for a dense ordering of pitch-classes."

Frequency classes involve ascriptions like ‘333 vibrations per second’, and pitch-classes are concerned with ascriptions like ‘A flat’, ‘C’ and so on. The ordering of frequencies can be dense, because there can always be subdivisions producing the result that between any two is another more like each of them than they are like each other. But this is not true for pitch-classes, and yet Goodman has suggested that they are also densely ordered in this way. According to Raffman, Goodman has specified a dense ordering of pitches by appeal to their perceptual relations.

Some of the supplementary markings refer to pitch, but cannot be classified by any means which depend on a normal human faculty. An instruction concerning timbre, for instance, may indicate a certain frequency, but we could never measure the compliance of a note to particular frequencies without some measuring instrument. On the other hand listeners can identify sound events as compliant with markings in the score. Thus according to Raffman, the supplementary markings are semantically undifferentiated - we cannot determine (without some technical instrument) which supplementary pitch marking any given sound complies with. If the supplementary markings are semantically undifferentiated, then Goodman’s view that they are not part of the score because they are semantically differentiated, is severely threatened.

It would not follow that the supplementary markings of tempo and so on, are part of the notated score. But since we have seen how Goodman’s view can be modified to offer us an ideal for performance, rather than specialised criteria, we do not have to suppose that supplementary markings are part of the score in order to give them the intentional power and relevance that Ingarden suggests they have. if Goodman’s claim is weakened by Raffman; and modified by Goehr, then the intuitive belief that some performers hold that performance involves reference to the composer’s instructions as set forth on the score can still be held.

Such supplementary information may not be notational, but it has been suggested that scores are not entirely notational, and that a view that claims they are, has misconceived the compliance relation that holds for such markings. With Goehr and Raffman on the side of the authentic performer, there does not appear to any reason why he or she should not take the score to be identified with the music; in respect of the notated pitches and the supplementary verbal markings. Practising musicians have always taken this to be the case, pace Goodman’s claim that to do so is a disaster for theory. But that they do so is not as disastrous as Goodman thinks, because he has ignored questions of the intentionality of the score; and has not treated the notation and supplementary instructions as complimentarily presenting an ideal towards which performers strive.

1.4 The intentionality of the artwork

If we hold that the work is in some way identified by its score, and that the verbal instructions of the composer are constitutive parts of the score - i.e. that they are of equal significance to a performer, then we must ask about that significance. For it may be assumed that the score is directive, in that it consists of instructions issued by the composer to the performer(s). Before we can ask - as I shall - whether such directions must be followed, we must show that such instructions are intentional. If we show that they are; then we can ask whether their intentionality demands of us that we follow them.

But first, let us consider the intentional status of a musical artwork. I am going to assume - because I see no reason to do otherwise - that works of music are artworks. Borderline cases such as the works of John Cage exist; and it may be said that Muzak is not art, even if it is music. But these issues do not concern us here, and I am speaking - as do the authenticists - of the pieces of music which we invariably come across in the Western tradition of the last four hundred years or so.

1.4.1 Artworks as artefacts

It us generally accepted that artworks must be produced by human beings. If nothing more can be said about the definition of ‘Art’ this much is clear, that originating from a human creator is a necessary condition for artworks. It is the human origin of artworks that helps to distinguish them from many other kinds of object. As Paul Valéry says:

"Originally the word Art meant simply way of doing. This unrestricted sense has gone out of use."

While this sense may have gone out of use, we still have the vestige of art-as-human product in our concept. Anthony Savile believes that this vestige at least is to be assumed:

"I doubt that anyone would deny that works of art are intentionally constructed artefacts, but it is easy to see how someone of a sceptical turn of mind might argue that this is of little significance. Art, he may say, is merely what is man-made and of aesthetic interest. This is the basis of the connection between art and intention, for what is man-made in standardly made intentionally."

Monroe Beardsley also acknowledges the rôle of intention in the creation of artistic aesthetic objects:

"The things that naturally come to mind when we think of works of art are the products of deliberate human activity, sometimes long and arduous ...these things were intended by someone, and no doubt they are largely what they were intended to be by those who made them.

The artist’s intention is a series of psychological states or events in his mind: what he wanted to do, how he imagined or projected the work before he began to make it, and while he was in the process of making it. Something was going on in Chaucer’s mind when he was planning The Canterbury Tales and in Beethoven’s mind when he was considering various possible melodies for the choral finale of his D Minor Symphony (No.9). And these happenings were no doubt among the factors which caused those works to come into being."

As Savile points out, a sceptic would be hard-pressed to deny that works of art are artefacts. There are, of course, artefacts which are not artworks, and these can be of aesthetic interest. A chemical compound may taste pleasant, or be nicely coloured; a plastic toy may be smooth to touch. Anything can be appreciated in virtue of some aesthetic quality which it has, but we should not say - as Savile’s sceptic might - that a work of art is merely an ‘aesthetically interesting’ artefact. If it were, then any artefact might be art, because anything can be aesthetically interesting:

"...one could economically dispense with the concept of art because all the work that it does can be carried out by a broader concept which we already employ, the concept object of aesthetic interest ."

Savile rightly points out that works of art cannot simply be artefacts, because we would then have no need of a concept of art. He continues with reference to the peculiarly artistic intention involved in the production of art:

"...the peculiar values of art... are dependent on the artist’s intention to give expression in his work to what he selects and presents for our appreciation, and it is to this intentional feature of the production of art that we may again appeal in showing why art cannot without loss sustain assimilation to the broader concept of object of aesthetic interest."

We need not concern ourselves here so much with what it is to ‘give expression’ to what is selected, but rather with the fact that it is selected. The artist makes choices, and while it may be true that he or she does not have complete control over what is presented (the artist cannot envisage every feature of his or her work); some knowledge and control is to be attributed, because the production of art is effected by intentional activity and is not a random process:

"The work of art is the true object to the formation of which the creative acts of the artist are directed... The work of art then, is the product of intentional activities of an artist."

1.4.2 Intentions

The writing of a score is an action of a composer, which involves the making of choices - either prior to writing, or at the time of writing. To the extent that a musical composition consists of series of pitches, these pitches are selected by the composer, from a large, but finite sample.

The element of choice is crucial because, as Brian O’Shaughnessy points out, all action (and we should say that the production of art amounts to an action) involves choice:

"Thus, it is not just a necessary condition of an event’s being actively done, that events of that kind should be susceptible to the causal influence of the intention. This actually functions as the decisive test of action (as Wittgenstein pointed out in the 1930’s). That is, not that the act be the expression of the powers of intentional self-determination, but that it can be. That is, that it be the type of item which it makes sense to ask someone to do. That is, do choosingly. That is, intentionally as a result of deliberation... I recognise that intention is located at the very heart of action... in so far as causal sensitivity to the intention is the decisive necessary condition of action. And, in any case, in that intentional action provides the rationale for action itself."

Intention is inexorably linked to action. Wittgenstein knew this (as O’Shaughnessy points out), and says:

"615. "Willing, if it is not to be a sort of wishing, must be the action itself. It cannot be allowed to stop anywhere short of the action."... But it is also trying, attempting, making an effort, - to speak, to write..."

But,

"it is not merely towards; that it is not merely towards and in fact coming-from; but that it is internally and in any case avowedly from-towards. "

According to O’Shaughnessy, intentions are future-oriented. Elizabeth Anscombe does not agree:

"... if you try to make being concerned with the future into a defining property of intentions, you can be asked what serves to distinguish this concern with the future from the predictive concern."

O’Shaughnessy would reply that the past influences our intentions toward the future, and even though we can sometimes be said to have an intention in the present sense, it is still grounded in a past-to-future construction:

"A homely example demonstrates this truth. Thus, mid-way through a sentence, I cannot be performing the intentional deed that engages me, unless I have, not merely some idea of what I shall say next , but some idea also of what I have just said!... the past enters internally into the intention as it does not with the unself-conscious. This shows in the fact that we bring human intentional acts under essential past-determined descriptions which continue to apply mid-way through the act, say ‘striving to swim from Dover to Calais’,"

The notion of present-tense intention is grounded in this:

"Thus, while it is true that the man swimming from Dover to Calais retains that intention even as he is mid-way through the act, the subordinate intention whose content gives the remainder of the intention yet to be expressed is exclusively directed future-wards ... the as yet unexpressed sector of the intention has to be exclusively directed towards an act-part whose beginning point is no earlier than the present."

We do not need to think about the notion of intention for very long before we realise that there are activities apart from the creation of art which are intentional (activities which result in the production of what we may call ‘intentional objects’). Works of art are artefacts, and artefacts are intentionally produced. This is perhaps where Savile’s account is inadequate, because there are many things which are ‘intentionally constructed artefacts’, which we do not think of as works of art, and if we contend, with Saville, that all artefacts are ‘standardly’ made as a consequence of intentional action, then the notion of an ‘intentionally constructed artefact’ is tautological. Being an artefact, and being intentionally constructed, are one and the same; and while artworks are both, neither description distinguishes them from other (intentionally constructed) artefacts.

H.L.A. Hart puts it differently, and reminds us of Bentham’s distinction between oblique and direct intention:

"Some legal theorists, Bentham among them, have recorded this divergence by distinguishing (as ‘oblique intention’) mere foresight of consequences from ‘direct intention’ where the consequences must have been contemplated by the accused not merely as a foreseen outcome, but as an end which he set out to achieve, or as a means to an end, and constituted at least part of his reason for doing what he did."

Anthony Kenny fleshes out this distinction, in such a way that makes it appropriate to this inquiry:

"Jeremy Bentham, followed by other philosophers, made a distinction between direct and oblique intention. Something was directly intentional if it was something sought as an end in itself, or as a means to an end; something was obliquely intentional if it was neither an end nor a means but merely a foreseen consequence of one’s directly intentional actions... If I get up in the middle of the night to get a drink of water, knowing it’s likely that I’ll wake the baby, but hoping very much that I won’t, then getting a drink and getting out of bed are directly intentional, while waking the baby (if the brat wakes after all) is only obliquely intentional."

When art is made, something is done such that it becomes appropriate to talk in terms of what the artist was trying to do, and of what he intended by doing it. As O’Shaughnessy reminds us, there is a distinction here:

"Thus, the intention is a something, generated by desire and belief, causing knowledge of itself, belief in the occurrence of its object, and any higher-order desires to do that deed, and tending to cause striving at the appointed time... the intention acquires its strangely mixed character: as sense-fixer of , as practical commitment towards... For we should never lose sight of the fact that we both mean to do certain acts, and mean something by or in doing them."

Where art is concerned this distinction is particularly pertinent. The artist sets out to do something e.g. paint a picture, and hopes that it will turn out to possess certain features. The intention to make a picture is direct, in that it is the end towards which the intentional activity of painting is directed. Whatever qualities the painting turns out to possess owe their existence to this intentional act of painting; yet we should not say that they are intentional in the same way that the painting itself is intentional. We may say that the qualities of the painting are obliquely intentional, in that they may have been foreseen by the artist, as consequences of the directly intended brush-stroke actions which were undertaken.

Some qualities of artworks are plainly not foreseen by the artist, even if they are observed after completion of a work. They may be foreseeable however, and qualities which damage a work, or which are detrimental to it, are often of this kind. Composers do not generally compose pieces with the intention that they be weak structurally, or that they be over-indulgent. There are pieces like this (many of which are no longer extant), and although the artist may have intended something which turned out to be detrimental to the work; he may also have intended to succeed in something in which he actually failed.

One of the things we can consider about a work is the extent to which it is successful:

"... one of the sources of our appreciation in contemplating a work of art is our awareness of the fitness of the work to the end for which it was produced. This... can only be appreciated, however, if the artist’s ulterior intentions are understood and borne in mind."

The extent to which the artist’s intentions can be successful is governed by his or her ability to carry out the actions which fulfil the intentions. We can turn to O’Shaughnessy’s notion of ‘power’ to help us here:

"A champion can rationally intend and a novice irrationally intend to hit the bulls-eye on a dart board; but while both can succeed in doing what they tried and intended doing, only the champion’s act counts as an intentional hitting of the target. Why so? At first glance the reason seems to be that the champion possesses, but the novice lacks, the power to do what he tried to do."

There is a problem with this, in that having the power to do something need not entail the ability to do it all the time. But neither does having been able to do it once or twice in the past mean that one has the power to succeed in doing it again and again. Having the power to do something need not involve the ability to do it at will. O’Shaughnessy says:

"The power exercised in intentional action must be such that the content of the causative intention was significantly causally contributory to the success of his trying... Now when the content is thus significantly contributory but not an absolutely sufficient condition, so that he cannot do the deed at will, then I think we must be dealing with a law that links trying and success in optimum display conditions, not categorically, but statistically significantly ... for we characterise acts as intentional only when the power that found intentional expression was of non-negligible extent."

This idea of something’s being significantly contributory is a little vague, but it does help us to distinguish between flukes and successes; and enables us to say that the success of certain intentional acts (such as in sports), are neither flukes, nor deliberate choices. A marksman does not choose to hit the target; neither is it a fluke if he does so. The marksman has a degree of skill, which affects the probability of an accurate shot. The same must surely be said for an artist - he or she does not choose to give a painting a particular quality, even if that is what is desired, but we do not fail to credit the artist if the painting subsequently possesses that quality.

Thus an artist who is skilled (and who knows how to exercise that skill), can intend to produce a painting which possess a particular (aesthetic) quality. The intention need not be oblique, because the aesthetic quality may be rationally sought as an end in itself.

There are other qualities which we attribute to a work which cannot have been foreseen by the artist. Any Oedipal qualities in Hamlet (in the Freudian sense) cannot have been foreseen by Shakespeare, and so we should not even say that they were obliquely intended. Oblique intention, it may be remembered, is sufficient to condemn someone for murder, since it still bears the agent’s responsibility. A mother is not condemned for the crimes of her middle-aged son, nor can Shakespeare be held responsible for any Oedipal quality which Hamlet displays. He is responsible in the sense that he wrote Hamlet, but we do not give him credit (or blame) for qualities which could not have been foreseen, and which therefore, could not have been intended in any sense of the word.

1.4.3 The score as intentional object

Having seen how artworks are intentional, we can now see that scores are representative of the composer’s intentions concerning how the piece was intended to be played. We have discussed the relationship between the ‘score’ and the ‘work’, and we can see that the score, as the product of intentional action (the action to produce a notated score) can manifest the intentions of the composer. These intentions are often direct intentions, such as where a composer asks for a particular instrument, or tempo.

There is not a clash between the indeterminacy of the score and the directness of the composer’s intention. It may be thought that a composer cannot directly intend something which he or she cannot stipulate exactly. This need not be true.

That the score manifests such intentions does not mean that we have to perform the music as the composer wanted it performed. Or does it? Why should we respect the composer’s intentions?

The answer to this kind of question need not be normative. It may be very helpful to appreciation to take into account the composer’s intentions, and the milieu in which the piece was composed, but it would not follow that we must therefore do so. At best we might propose that it is a valuable thing to do, and suggest why; but I doubt that there could be a conclusive argument as to why we must. So I shall make no extravagant claims for the rôle of intentions, but shall present a case as to why they should not be dismissed, as some writers have suggested they should.

1.4.4 The case against intentions

The next question I want to discuss asks why we should not involve ourselves in intentionalistic criticism of artworks. The most famous argument made against such criticism was put forward by Monroe C. Beardsley and W.K. Wimsatt, and so I take them to be representative of the view.

That an artist can create art intentionally has never been in doubt. That art is created is significant, in that it betrays the fact that art, as Valéry pointed out, is something done. But that the art is intentionally done need not mean that the ‘intentional features’ of the work (the intentions ‘behind’ it), should be taken into account when describing or evaluating a work. A failure or success of intention may explain why it possesses or lacks a certain quality, as Roland Barthes points out, but if we are interested in artworks, either as artworks, or as aesthetic objects, we are likely to be interested in the qualities themselves rather than how they got there - we are interested in the object rather than its creator:

"The explanation of a work is always sought in the man or woman who produced it, as if it were always in the end, through the more or less transparent allegory of the fiction, the voice of a single person, the author ‘confiding’ in us. It is the language that speaks, not the author..."

Beardsley would want to adhere to the idea that we should be concerned with the text, rather than the author:

"Now the goodness in which we take an interest (when our interest is aesthetic), is something that arises out of the ingredients of the poem itself: the ways its verbal parts - its structure and texture - combine and co-operate to make something fresh and novel emerge...

It is in its language that a poem happens. That is why the language is the object of our attention and of our study when its meaning is difficult to understand. It is not the interpreter’s proper task, then (I argue), to draw our attention to the psychological states of the author."

We should acknowledge that the author has had an influence on his or her work. To suggest that an artwork does or did not have a creator is absurd - and has not been claimed anyway. So we have not doubted that the work has a creator, but according to Beardsley and Wimsatt, the decisions and choices made by that person are not relevant to any criticism of today. Their approach to the issue yielded the claim that the intentions of the author (composer) are neither (1) available, nor (2) desirable:

"... the design or intention of the author is neither available nor desirable as a standard for judging the success of a work of literary art."

1) Firstly, they are not available, because:

"There is a gross body of life, of sensory and mental experience, which lies behind and in some sense causes every poem but can never be and need not be known in the verbal and hence intellectual composition which is the poem."

That body of life, as far as Beardsley and Wimsatt are concerned includes what the composer intended the work to be like; how he or she wanted it to be viewed/performed; and the psychological ‘personal’ influences which affected the creator:

"I propose to count as characteristics of an aesthetic object no characteristics of its presentations that depend upon knowledge of their causal conditions, whether physical or psychological."

These causal conditions are not part of the work, so they cannot be detected by looking at the work, neither can they be determined in any other way.

"The poem is not the critic’s own and not the author’s (it is detached from the author at birth and goes about the world beyond his power to intend about it or control it)."

The author cannot claim any special position of interpretative importance merely because he or she is the creator - the work ‘leads its own life’ so to speak. Thus, even access to a still living author would not make these features any more available. If such an author did give some information, we should be aware if two possible responses Beardsley and Wimsatt might give.

Firstly, they might refer to Beardsley’s dictum that such features are not part of the work. The work, Beardsley has said, must be distinguished from the intentions which created it:

"...we must distinguish between the aesthetic object and the intention in the mind of its creator."

Even if the object turns out to be exactly what the artist intended it to be, we must still distinguish between those intentions, and the finished artwork, which has an independent existence.

Secondly, they might reply that the author, if he or she does give some information, might have a vested interest in doing so, and so might not tell the truth; or might not remember accurately what prompted the production of the work, or what he or she had meant by it. Their argument might thus be that just as we do not have access to information about the psychological causes, nor does the composer after the fact. The living artist’s responses are just as (un)likely to be accurate in this respect as certain other people’s. We should not be tempted into the position of asking the author, under the assumption that he or she will be in a position to know the answers. Beardsley and Wimsatt say that he or she could not know the answers.

Thus, according to Beardsley and Wimsatt, the artist’s intentions are not available, because the artwork (which he calls ‘aesthetic object’) which we encounter leads a life of its own, apart from its creator. Also, since the artwork is to be distinguished from those intentions, it must necessarily be the case that questions concerning them are not questions concerning the work.

2) Part of the reason that information about the author’s intentions is not desirable is because such information does not concern the work. Information about the author is not information about the work, and since we should only be interested in the work itself, we should not be interested in features which are not part of it. Thus, even if we showed that some information was available, they would still wish to deny us to opportunity to base our critical comments upon the artist’s intentions. They would still say that even those which were available were not desirable:

"Nevertheless, we submit that this is the true and objective way of criticism, as contrasted to what the very uncertainty of exegesis might tempt a second kind of critic to undertake: (2) the biographical or genetic enquiry, in which, taking advantage of the fact that Eliot is still alive, and in the spirit of a man who wants to settle a bet, the critic writes to Eliot and asks what he meant... Our point is that an answer to such an inquiry would have nothing to do with the poem ‘Prufrock’; it would not be a critical inquiry. Critical inquiries, unlike bets, are not settled in this way."

It is surprising that Beardsley and Wimsatt conclude their paper on this note, in the light of a distinction at which they hint, but which is crucial:

"The use of biographical evidence need not involve intentionalism, because while it may be evidence of what the author intended, it may also be evidence of the meaning of his words and the dramatic character of his utterances."

If the inquiry after biographical or genetic information is inappropriate, and it is distinguishable from inquiry after intentions, then it would not follow that inquiry after intentions is inappropriate. Thus we might finish reading a paper which either confuses the issue or which is not really about what its title claims it is about. It seems that Beardsley and Wimsatt are really trying to eliminate the possibilities for ‘sentimental’ criticism. There is a difference between saying that the artist’s intentions are irrelevant, and saying that one should not let what one knows about the author prejudice one’s judgement of the work.

Furthermore, if intentions are unavailable, there could hardly be any reason for desiring them; and if they are undesirable, then it hardly matters whether they are available or not. If they cannot be known, then there is no question as to whether they need to be known. The idea of a work’s ‘success’ is telling, and we might wonder what there is which can be successful. What does a work of art strive to achieve, yet which the artist does not strive to achieve? It may be said that any reference to the ‘success’ of the work carries the author and his intentions along with it. The artist has the intentions, not the work, and it makes more sense to think of the artist as trying to achieve something through the activity of creating art. The successful realisation of intention is not sufficient to make a work a good one, even if the artist generally has the power to fulfil such intentions regularly.

It might be better to say, not that intentions are available, but that intentions ought not be considered as unavailable. Reference to intentions is so difficult and unreliable that we should not do it:

"the three types of evidence... shade into one another so subtly that it is not always easy to draw a line between examples, and hence arises the difficulty for criticism."

The intentions are not available when they cannot always be distinguished from other evidence we might wish to draw upon (such as biographical and genetic information). But it does not follow that there cannot be cases where evidence concerning the intention is available.

1.4.5 The availability of the artist’s intentions

If we were to ask the question, ‘Are intentions ever unavailable?’, we would expect the answer to be ‘yes’, since we could probably think of some work of art about which we knew nothing. However, if we know that something is a work of art, we know - if what I have already said is true - that it is intentionally made. But there might be an intentional object, about the nature of which we do not know, and so it is conceivable that there could be an object and that we could not tell whether it was man-made (intentional), or natural. For such an object, the intentions would be unavailable, but only for as long as we did not know that it was an intentional object.

However, if we know that an object is an artwork, the intentions cannot be completely unavailable, because we know that it has been offered for someone’s (and perhaps not our) contemplation. To ask if intentions are ever unavailable is thus rather a fruitless question, because knowledge of intention becomes available as soon as we recognise something to be art.

Intentions may be thought to be unavailable in that information which Beardsley and Wimsatt would call ‘external’ may be lacking. We may not be able to write to an author and ask him or her what he or she meant, or read memoirs. Even if we could, they say, we should not, because:

"Critical inquiries are not settled by consulting the oracle."

If we do, we are consulting what they call ‘external’ evidence of intention, which addresses itself to some object other than the work of art under consideration. But we can consult such information, which, as Beardsley and Wimsatt acknowledge, can give ‘evidence’ of intention. I think it is a mistake to say that evidence of intention comes only from some such other object. If such evidence could only come from ‘outside’ the work, then Beardsley and Wimsatt’s claim might be true. But I suspect that not only can an object display an intentional origin (in virtue of its being a work of art), it can also reveal something about the nature of those intentions.

1.4.6 Personal qualities

Colin Lyas has written about what he calls personal qualities which are attributed to artworks and which presuppose the existence of the artist. These features, he claims, entail reference to the artist’s intentions. They also serve to distinguish artworks from natural objects:

"6. I offer now a list of merit and demerit qualities of works of art, qualities special to them as man-made things. We may, first, praise a work of art by using some of the following of it:

responsible, mature, intelligent, sensitive, perceptive, discriminating, witty, poised, precise, self-aware, ironic, controlled, courageous.

Demerit terms include the following:

simple-minded, shallow, diffuse, vulgar, immature, self-indulgent, uncomprehending, heavy-handed, gauche, glib, smug.

These terms characterise works of art but not natural objects. I call them personal qualities..."

For an example of the use of personal qualities, we may turn to Leavis commenting on George Eliot’s The Mill on the Floss:

"To understand immaturity would be to ‘place’ it, with however subtle an implication, by relating it to mature experience. But when George Eliot touches on these given intensities of Maggie’s inner life, the vibration comes directly and simply from the novelist, precluding the presence of a maturer intelligence than Maggie’s own. It is in these places that we are most likely to make with conscious critical intent the comment that in George Eliot’s presentment of Maggie there is an element of self-idealization. The criticism sharpens itself when we say that with the self-idealization there goes an element of self-pity. George Eliot’s attitude to her own immaturity as represented by Maggie is the reverse of a mature one."

Such terms as ‘mature’ are applied to works and are not merely descriptions of a separate author. Rather they are descriptions of the ‘author-in-the-work’. Eliot’s own immaturity shines through the work to such an extent that it is difficult to distinguish between Maggie’s and Eliot’s immaturity. Since Eliot herself is immature, she can neither create a mature character, nor display a mature attitude to such a character. Hence Eliot’s characterisation of Maggie Tulliver is transparent, in that it reveals her own immature disposition. The text then, can tell us about the author. The work may possess qualities which betray features of its creator’s personality.

Lyas says that these terms presuppose an author, although Joseph Margolis does not quite agree:

"Certainly this sometimes obtains - but not always. It shows the aesthetic relevance of biographical and intentional considerations; but sometimes an inspection of a given work supports their ascription - from which, therefore, we may draw inferences regarding the artist."

Thus it seems that information about the author enables us to attribute personal qualities to a work; but we can also detect such qualities without knowing anything about the artist, which amounts to saying that we can learn about Eliot and what she was trying to do in or by the work by looking at it. Thus we do not need to acquiesce to Beardsley and Wimsatt’s criticism that the artist and work are distinct, and that knowledge of the work does not involve reference to the artist. For we can learn about the artist of the work, by discovering the artist in the work.

There are examples for music too. Margolis quotes Guy Sircello on Peter and the Wolf:

"Prokoviev’s Grandfather theme is witty because the composer wittily comments on the character in his ballet"

The wit is supervenient upon the wit of the composer. For it is not possible for a composer who lacks wit to write a witty piece. The human qualities can thus be distinguished from emotional qualities in this respect, since we do not claim that a composer must be grieving in order to write music which is expressive of grief. The creator must own the personal quality himself though, even if he does not always write witty stories. But a witty story must have a witty author.

We can conceive of someone who has a disposition to be witty telling a story which lacks wit, or conversely, of someone who is not generally witty, being so occasionally. We can also conceive of someone who has a disposition to be witty telling a witty story, and when this happens we are reminded of O’Shaughnessy’s claim that an intentional act is an exercise of a power that is significantly contributory to the achievement of the intention. We may say that the people who have dispositions to be witty have the power to intend the products of their intention to be witty.

Someone who is mature and sensitive can produce a work which is immature. Where this happens, we do not want to say that he or she is demonstrating her own immaturity, because an immature work need not be authored by an person who is immature - the author may imitate or pretend an immaturity. An author who has the power to write sensitively may also have the power to write gauche or glib literature. A sensitive creator can intend to produce glib writing in a way in which someone who is generally heavy-handed cannot intend to produce controlled writing, and thus it must follow that we should not assume that the author of a smug text, and the composer of a vulgar tune are smug and vulgar themselves, while yet maintaining that the author-in-the-work is so.

In Shostakovich’s Ninth Symphony, for instance, there is a horn theme which may be described as vulgar. Should we say that Shostakovich was a vulgar man then? We know, from listening to his other music, that such a gesture is uncharacteristic, and from his writings and those who knew him, that he was a rather sensitive man. The two pieces of information we have do not tally until we consider another idea - that Shostakovich got into severe difficulties with Stalin over the kind of music he was supposed to write. We might remember, for instance, the history of the Fourth and Fifth symphonies, the former of which he suppressed because of Stalin’s condemnation of his opera Lady Macbeth of Mtensk District in 1936. The Fifth Symphony, of course, was written as ‘A Soviet artist’s reply to just criticism’, after this condemnation. Now Beardsley and Wimsatt, among others would say that this information is irrelevant to criticism, but it points us towards an interpretation of the Ninth Symphony which can be useful.

Stalin wanted Shostakovich to write a symphony which would celebrate the victorious conclusion of the war, and which would also demonstrate to the world that the Soviet Union had a great composer worthy of the throne of Beethoven. The ninth symphony is, by tradition, great, Schubert, Beethoven, Bruckner and Mahler had demonstrated this. Shostakovich’s ‘response’ (and it seems strange not to think of the composer as unaware of these expectations), was to write:

"... one of the shortest and, unquestionably, the wittiest and most cheerful of (his) symphonies."

Not only has the symphony been described in terms of wit, some evidence is presented for Shostakovich’s intending it to be so:

"This feature of the composer’s personality is reflected naturally in his works. ‘When the public laughs or simply smiles during the performance of my works, this gives me pleasure’ he said. These words may be applied with complete justification to his early ballets, to the first concerto for piano and orchestra... and to the Ninth Symphony."

Thus it seems reasonable to say that Shostakovich intended the horn motif in the third movement to be vulgar, because there is reason to suppose that he intended the symphony to be cheerful and witty, and he had the musical ability (power) to do this. Given that he also did not like the dictatorship of Stalin, who even went so far as to tell him what and how to write, we might suppose that some of the comic gestures are directed at Stalin. The horn theme might thus be interpreted as a vulgar gesture, made at Stalin. The work as a whole might then be seen as some kind of defiant gesture to the Soviet musical establishment, in virtue of the symphony’s flagrant disregard for the tradition to which it cannot fail to relate.

As well as being an example of ‘intentionalistic’ criticism, the foregoing contains references to personal qualities which betray the intentions of the composer. It is Lyas’ view that by referring to personal qualities one drags along the intentions of the artist, in which case it will be hard to say that they are irrelevant, because to do so would require a repudiation of such qualities. Use of personal terms implies a creator, in virtue of the fact that natural objects cannot have these qualities attributed to them. To call a sunset ‘witty’ is to treat it as though it was man-made. Personal terms are generally confined to the description of artefacts.

The claim then, is that there is a class of descriptive terms which we can apply to certain aesthetic objects, which are dependent upon those objects having a human creator. The creator can intend his work to manifest such qualities if he or she has the power to do so, in which case the qualities, when detected, demonstrate not only that the work has a creator, but what kind of qualities he intended it to have.

A creator cannot produce a work possessing a personal quality which he or she does not possess him/herself. He or she can produce a work possessing a quality which is not so sophisticated - such as a quality which can be easily pretended by that creator. Thus the qualities which an creator can produce, or can intend to produce, are supervenient upon the qualities which the creator possesses him or herself.

In response to Lyas’ claims, Beardsley and Wimsatt might want to argue that he has not construed the concept of the artist’s intention in the same way as them. They might have admitted, for instance, that Lyas’ personal qualities are in fact not ‘external’; but suggested that they are not ‘intentional’. Lyas talks of the artist-in-the-work, as distinct form the external artist, and Beardsley and Wimsatt may not object, because Lyas has still not done what they would have wanted him not to do - he has not involved external features in his criticism of the work.

Of course, Lyas would probably reply that his personal qualities are inexorably linked to ‘external’ considerations concerning the lifestyle, personality and cultural milieu of the artist, who cannot be completely separated from the artist-in-the-work.

1.4.7 The desirability of the artist’s intentions

There is a sense then in which we may say that the intentions of the artist in the work are available. To what extent are they also desirable? In this section I want to propose that there is a good reason for our being interested in the composer’s intentions, which serves both to explain why we are interested; and also offers a justification for being so. However, I am not going to suggest that we must look to the artist’s intentions, only that it is reasonable and beneficial to do so.

1.4.7.1 The relevance of intention

Comments about the advantages of certain approaches to artworks, ultimately rest on a question about relevance. For as soon as we start to defend approaches to artworks, or to proscribe what may or may be taken into account when we make judgements, we find that we are stipulating what is relevant to judgement. Mary Mothersill spends several pages of her book dealing with questions about relevance. It is often said that intentions are not relevant to criticism of artworks, but this begs a question, in that there are different kinds of criticism, one form of which may be characterised as ‘intentionalistic’ criticism. Intentions must be relevant to that. So, with Mothersill we might wonder:

"What sort of question is the question of relevance? Beardsley takes it to be conceptual; for Aiken it is empirical - what someone finds relevant is relevant."

Starting from Beardsley and Wimsatt, we have seen that there is a tendency among some theorists to say that certain features are (or are not) relevant to criticism. If there is a distinction between an artistic object and an aesthetic object (and there must be if there is a difference between art and nature), we should ask ‘relevant to what?’. Mothersill begins her work with an acknowledgement of the importance of this concept:

"But consider how many of the familiar problems of aesthetics prove to turn on what I shall call ‘the question of relevance’, namely the question of what considerations have or ought to have a bearing on the interpretation and valuation of works of art."

She is interested she says, in works of art, and so she might say that Beardsley has a different project in that a project ‘sets the stage for questions of relevance’. We might not expect the same features to be relevant to criticism of aesthetic objects as for artworks:

"A ‘consideration’ is relevant if it supports a critical verdict, and to do that it must involve reference to a property that has proved in the past to be a virtue, that is, to contribute to the goodness of artistic merit of a work of art. Relevance is thereby tied to principles of taste and principles of taste to laws of taste... and if... there are neither laws nor principles, then the question of relevance will have to be recast or else abandoned.

She does in fact recast the notion of relevance, and does not maintain the view that artistic criticism is a project. Critics may set themselves a project, and may thereby establish conditions of relevance, but these conditions are arbitrary according to Mothersill:

"There are rules of thumb, but no principles: what works in one case for one listener may draw a blank in different circumstances. ‘Principles of relevance’ such as that which excludes reference to ‘author psychology’ or general background information, go by the board. That Milton was blind, that Henry Moore has a collection of beach pebbles and driftwood, are thoughts that might enliven perception - thoughts rather than information, because for the critical enterprise, it doesn’t matter whether what is alleged is true."

Mothersill’s view is persuasive, in that it accounts for all kinds of critical methods which we might adopt. Freudian interpretation is acceptable, as is interpretation based upon what the author says about his work. Each may be a different project, and the critic sets up his or her own project. Consequently, the notion of relevance is itself irrelevant to the attempt to determine the ways in which we appreciate artworks.

If what Mothersill says is correct, then we cannot expect to find an argument that convinces us that we must involve ourselves in intentionalistic criticism. Nor are we likely to find one which necessarily prevents us from referring to the artist’s intentions.

1.4.7.2 The artist’s motivations

The intentions of the artist can be direct. Sometimes we want to know the artist’s reasons for doing something, whether we are interested in why a particular chord is inverted; or in why Mozart’s Requiem was written in the first place. When we ask these kinds of questions, we put the artist on trial, so to speak. We have in front of us the product of his or her intentional actions, and when judging it, we want to be in a position to make a fair judgement. It seems reasonable to seek evidence, not only in the product of the action, but in the action which funded it. It need not follow from this that we give any special credence to what the artist says, although it is reasonable to listen to what he has to say, both within and without the work. We ask someone accused of a crime to give an account of himself, but we are in no way obliged to believe him - we listen to what he says, and see if it tallies with the facts (or object) before us. This is like the attitude of Aiken, who told us to ignore the intentions if they did not coincide with the meaning of the work. There is a slight difference though, in that we are dealing now with what the artist says about the work, rather than what the work says about him or her.

Now it may be said that in dealing specifically with direct intention, we are concerning ourselves with the artist’s motivations. Motives for a crime are significant, although no-one may be convicted on the evidence of motivation alone. It is not only reasonable, but useful (it is reasonable because it is useful) to ask someone why they committed a murder, and the same seems to be true where artworks are concerned.

Enquiring after motivations can help us to understand an action, and this is inexorably linked to an understanding of the consequences of the action. A motivation, in law, can be interpreted as a kind of cause - "Othello killed Desdemona because he believed she was unfaithful"; and similarly (pace Beardsley) in artistic criticism, "Berlioz used a diminished seventh chord to open the Songe d’une nuit du sabbat because he wanted to conjure up associations of fear". We not only can make statements like this about artworks - we often do:

"Nothing could be commoner among critics of art than to ask why the thing is as it is, and is characteristically to put the question, for example, in the form ‘Why does Shakespeare follow the murder of Duncan with a scene which begins with the sound of knocking?’ The best critic is the one who knows best where to ask this question, and how to get an answer; but surely he doesn’t feel it necessary, or desirable even were it possible, to get in touch with the artist to find out the answer."

Many people find it helpful to ask questions such as Cavell’s, presumably because they feel that the knowledge gained makes the artwork valuable in a particular way. The question we are asking here, is, "Why did the artist do that?", where "that" refers to a particular feature or combination of features of the work. This kind of question is not always asked directly, such that it obviously refers to the artist. "Why did the artist do that?" also manifests itself in the form of questions about the work - "Why does Parsifal baptise Kundry" translates into "Why did Wagner have Kundry baptised by Parsifal?". A question about the work may be taken as a question about the artist’s motivations for carrying out a particular intentional action, such as the stipulation of certain actions in a performance; or the description of a particular fictional event, or emotion. "Why does Othello kill Desdemona?" can also be construed as "Why does Shakespeare make Othello kill Desdemona?". Answers referring to the ‘given-ness’ of the story would only be trivially true.

Questions about motivations, (which distinguish the ‘best’ critics, according to Cavell) can be fairly simple, or complicated. If we ask why Berlioz used a certain chord, the answer may be straightforward and brief. But if we ask why John Cage wrote 4’33", we would have to investigate details of his personal character and historical and cultural surroundings, as well as read his writings. I asked this question when I discussed Shostakovich’s Ninth Symphony, and the asking indicated what kind of information was useful. That we can ask these questions is not in doubt. What is of interest though, is why we should want to know the answers, and to what use we can put them.

1.4.7.3 The nature of our relationship to artworks

Mary Mothersill quotes Cavell making what she calls a ‘good point’:

"... the answer to the question ‘What is art?’ will in part be an answer which explains why it is we treat certain objects, or how we can treat certain objects, in ways normally reserved for treating persons."

It seems to me that we treat artworks and people in similar ways. There is, I think, a useful analogy to be drawn between ‘liking’ artworks, and liking other people.

Our love of art can be both particular and general, as can our love of people. We like certain works of art, and we like art. Some people just are not ‘into’ art - which is their loss of course, and there is no accounting for taste. On the other hand, some of us enjoy the company of other people more than others, and we each have our own friends and lovers, who are not always liked by everybody else.

The acquaintance we make of a work can be startling and violent, or it can be a gradual familiarisation. Some people are appealing when we first meet them, others have to ‘grow on us’, so to speak. We do not keep our ‘artistic’ friends all our lives - we fall out of love as much as in love - as David Hume pointed out:

"A young man, whose passions are warm, will be more sensibly touched with amorous and tender images, than a man more advanced in years, who takes pleasure in wise, philosophical reflections concerning the conduct of life and moderation of the passions. At twenty, Ovid may be the favourite author; Horace at forty; and perhaps Tacitus at fifty... We choose our favourite author as we do our friend, from a conformity of humour and disposition."

The ‘friendship’ may last a very short time indeed:

"Not to mention, that there is a species of beauty, which, as it is florid and superficial, pleases at first; but being found incompatible with a just expression either of reason or passion, soon palls upon the taste, and is then rejected with disdain, at least rated at much lower value."

When we find something out about someone, such as their opinions, or family history, we often alter our judgement. We can like or dislike a work of art without knowing anything about it, but when we learn something about it, our judgement can be affected. The question of relevance becomes less useful when we consider what we like in art, because whether theorists like it or not, some people do not like Wagner’s music because they suspect him of anti-semitism, which they abhor because it reminds them of the Holocaust. Such beliefs may be personal associations (held commonly by large numbers of people even), which have nothing to do with the music, but they still exist, and affect individuals’ relationships with artworks and other people. We might just as well dislike the person of Wagner for the same reason. When we let our judgement be affected in this way (and we may not be able to help it), we find that we are not only responding to the music as we would a person, we are responding to it as though it were the utterance of the particular person who is the composer. Thus there is a possibility of confusion between treating the work as though it were a person; and treating it as though it is the person of the artist. I am more concerned with the former, since the latter is prone to the difficulty raised by Roger Scruton:

"I have argued that a minimum requirement of critical objectivity is that the critic should be able to draw the distinction between meaning and association - between what is in the work and what is not. (The sadness is in the music, my melancholy on hearing it is not.) It is surely not implausible to suggest that part of what enables us to make this distinction lies in a ‘sense of intention’ with which every work is imbued."

The qualities of the artist, of the work, and of the observer or listener, are to be distinguished. Sometimes we do not distinguish, and treat the work as though it were the person of the author, at least in as much as we allow knowledge of his or her personality to affect our judgement such that we judge the artist rather than the work. I think that it is also possible to talk about art in similar ways to the ways in which we talk about other people without going so far as to suggest that the work of art is no more than an embodiment of the artist’s opinions. The relationships we form with works of art are like the relationships we form with other people; they are not elliptical relationships formed with the work’s creator.

A lover desires to know and understand the beloved, which involves finding out about the life history, opinions and formative influences on that person. It is not that this kind of information is of interest that is important, but that the history of a person determines what she is, and so any desire to know a person will eventually lead to some kind of enquiry after the things which make her unique or special. The sharing of past history builds relationships, as Othello well knows:

 

This to hear

Would Desdemona seriously incline;

But still the house affairs would draw her thence;

Which ever as she could with haste dispatch,

She’d come again, and with a greedy ear

Devour up my discourse. Which I observing,

Took once a pliant hour, and found good means

To draw from her a prayer of earnest heart

That I would all my pilgrimage dilate,

Whereof by parcels she had something heard,

But not intentively. I did consent,

And often did beguile her of her tears,

When I did so speak of some distressful stroke

That my youth suffered. My story being done,

She gave me for my pains a world of sighs;

She swore, in faith, ‘twas strange, ‘twas passing strange;

‘Twas pitiful, ‘twas wondrous pitiful.

She wish’d she had not heard it; yet she wish’d

That heaven had made her such a man. She thank’d me;

And bade me, if I had a friend that lov’d her,

I should but teach him how to tell my story,

And that would woo her. Upon this hint I spake;

She lov’d me for the dangers I had passed;

And I loved her that she did pity them.

I am suggesting that the fact that we want to know so much about our friends and lovers helps to explain why we want to know so much about the works of art we like, and art in general. We do not infer the character of someone by merely looking at his or her face, we engage in conversation, and to do so no more goes ‘outside’ the person than to enquire after the genesis of a work of art goes ‘outside’ it. We can confine ourselves to the aesthetic features of either the person or the work, but to do so is to miss out on the full story, and may even be unfair - particularly where a person is involved. Beardsley and Wimsatt want us to ignore such considerations (they want us to go by appearances, so to speak), yet they acknowledge that we do sometimes investigate the genesis of the work, although they do not offer any explanation as to why we should want to do this. We may be led, naturally and reasonably, to a knowledge that it was made under certain conditions - such as a fear of Stalinist purges - in order to present some view or attitude. It might be that what is discovered becomes a reason for qualifying our judgement, or for admiring the work the more. We are interested in these issues because we are interested in the work: to be so is relevant to the project in mind (the full appreciation of the work, in the sense that when we appreciate a person we investigate these issues).

This enterprise is rather like what Richard Wollheim calls ‘criticism as retrieval’:

"The task of criticism is the reconstruction of the creative process, where the creative process must in turn be thought of as something not stopping short of but terminating on, the work itself. The creative process reconstructed, or retrieval complete, the work is open to understanding."

In seeking to find out all we can about a work, we must investigate the creative processes of the artist, and in doing so we should expect to include some understanding of his or her intentions. Wollheim goes on to suggest the objection that his notion may be impossible to put into practice, and that some notion of scrutiny may be better employed. However, scrutiny leaves something out:

"Suppose we confine ourselves... to that part of the creative process which is realized in the work of art. It becomes clear that there is something that reconstruction of this part of the process can bring to light which scrutiny of the corresponding part of the work cannot. It can show that the part of the work which came about through design did indeed come about through design and not through accident or error. Scrutiny, which ex hypothesi limits itself to the outcome cannot show this. ... scrutiny, to be a source of knowledge, must presuppose retrieval."

Wollheim is in favour of referring to beliefs which could not be gleaned from a mere perception of the work, since he believes that they contribute to the understanding of the work. Some information concerning the genesis of the work gained from sources other than the work (such as the author’s autobiography) are part of the creative process, and can be useful:

"In other words, reconstruction of the creative process is admitted as the, or at least a, central task of criticism, but it must have a purpose in mind, and that purpose is that its findings should be put to use in scrutinising the work."

Under Wollheim’s account, intentions are relevant, as part of the ‘creative process’, but he also includes the ‘vicissitudes’ to which the human creator is subject:

"Some of these will be themselves intentional - change of mind - but some will be chance or uncontrolled. Secondly, the creative process includes the many background beliefs, conventions and modes of artistic production against which the artist forms his intentions: amongst which will be current aesthetic norms, innovations in the medium, rules of decorum, ideological of scientific world-pictures, current systems of symbolism or prosody, physiognomic conventions, and the state of the tradition."

Wollheim’s first statement needs qualifying in respect of what I have said about intention - certainly there can be factors which arise in art owing to chance, or lack of control - but we have seen how these might in fact be intentional in an oblique sense. I doubt that Wollheim would want to exclude such factors from his creative process though, since he tends to include features which are both the deliberate and accidental consequences of the artist’s intentions. Sometimes such ‘chance’ or ‘uncontrolled’ features may also come under this description.

Wollheim’s account is somewhat vague though, in that if I were to adopt it I would have to accept that the consideration of almost anything could be relevant as an ‘artistic’ judgement. This is unsatisfactory, if only because it makes an ‘anti-semitic’ interpretation of Parsifal artistic, when all that such an interpretation does is take some extraneous fact about Wagner and treat it as relevant. When we ask "Why did the artist do that?", we expect to get an answer which is convincing, and useful. To say, for instance, that Klingsor, being the villain in the work, must be a Jew because Wagner did not like Jews, is weak, and so any judgement which follows from that is weaker still. Kundry may well be Jewish (through having been the mythological woman who spat at Christ), but this is hardly ‘relevant’, because it does not help us to understand the work with which we want to form a relationship. Similarly we do not want to hear too much malicious gossip about our friends, who do not become our enemies through keeping bad company. We do not need to know that Wagner (or Kundry for that matter) is anti-semitic, because that knowledge is not useful.

Thus there are three points to be made about the desirability of information about the artist’s intentions.

1) Firstly, following Mothersill’s view - the question of relevance cannot be answered in such a way that prevents us from being interested in the composer’s intentions. Certain kinds of criticism benefit immensely from reference to such intent, while others, no doubt, do not.

2) Also, we often ask questions of artworks, such as "why did the artist do that?" in which case we are asking after the artist’s reasons and motivations. This is something which even those who eschew intentionalistic criticism indulge in, by virtue of the intentionalistic nature of artworks, very many questions translate into questions which enquire after the artist’s motivations.

3) Finally, it may be the case that we are attracted to artworks in a similar way to the way in which we are attracted to other people, and so our interest in the intentions, and background to a work is accounted for rather in the same way as our interest in the equivalent information we seek for other people in whom we are interested. We do not generally predetermine what we want to know about out friends, nor do we where artworks are concerned - unless of course, we have some ideological or philosophical bias to exercise.

1.4.8 Performance as interpretation

We have been talking about the role of intentions in criticism. I want to raise the question as to whether what Beardsley and Wimsatt say applies equally to music (as representative of a notated performance-art). They are discussing criticism and it may not follow that their comments on intention apply equally to performance. Beardsley and Wimsatt direct their complaints against critical interpretative use of intentions. But performance is a different kind of activity.

It is important to distinguish between interpretation and performance, particularly as we often talk of performers as interpreters. Jerold Levinson and Göran Hermerén tell us that performance is a form of interpretation, but it must be distinguished from verbal, critical interpretation. There is a sense in which we call a performance an interpretation. There is also a sense in which we call a description, or other verbal account of a work of music an interpretation. These two phenomena are to be distinguished, because while they may overlap, they are fundamentally different.

Levinson puts forward a distinction, introducing the terms critical interpretation (CI) and performative interpretation (PI). A PI, is said to be:

"... a considered way of playing a piece of music, involving highly specific determinations of all the defining features of the piece as given by the score and its associated conventions of reading."

A PI is not a particular performance, rather it is a way of performing, which can therefore be instantiated on several occasions. A PI, it seems, is a type, of which there may be locatably distinct tokens. A PI is intentional and deliberate, and its result is the production of particular sounds. There is little scope for ambiguity in PI, it is either not possible; or the result of attempting to present differing kinds of PI within a performance of set of performances is disastrous. For example, there has been some debate about the theme of the first movement of Mozart’s Piano Sonata in A, K331. Some performers phrase according to the natural division of the bars:

Ex 1.

......................

......................

There is a more common practice which suggests that the anacrusis is significant, whereby the phrasing is as follows:

Ex 2.

......................

......................

In writing about the work (CI), it is relatively easy to indicate a critical debate, but any attempt to introduce this issue in performance without making a commitment to one practice or the other would not be appropriate. Critical interpretations can cope with ambiguity and contradictory views, but PIs cannot. PIs, as distinguished from CIs, are less fluid - a PI is fixed, not in a particular performance, but in the manner in which particular performances are carried out. A performer could render the piece twice, with different phrasing - either at different concerts, or even in a lecture environment, but then he or she would be presenting two different PIs. Where a piece has equivocal character, a given PI must settle on one side, because it is not possible to do something in two different ways at the same time.

That a performer can make a critical interpretation (CI) of a work is not in doubt. This being so, we have to ask about the relation between a PI and a CI, because it is reasonable to suppose that a PI may be influenced by a CI. Even from my example above, we can see that a pianist may be influenced by the Shenkerian view that there is no anacrusis, and will thus construct a PI which is based on ex. 1. So for Levinson:

"What is truly implicit in Fleischer and Szell’s PI - a set of realizational choices - is at most a view as to how the work should be played. And a view as to how a work should be played is not thereby a view of the work, in the sense of a critical interpretation, though again, in many cases the performer’s view of how to play will indeed issue from and reflect his view of the work, its content and form."

But while a PI may issue from a CI, various CIs may be appropriate descriptions of a PI. There is generally a number of CIs which are compatible with a particular PI. Also, a performance may not instantiate a PI. A PI must represent a set of choices made by the performing interpreter, and some awareness of the options available. A PI is not merely a recognisable performance, it must be interpretative in some way, and there is more to interpretation than performance. Levinson’s implication, is that interpretation involves the making of choices - the taking of decisions where the score demands them. This must mean that a performance that instantiates a PI may be aurally indistinguishable from one which does not. We may think of our comments on intention here though, for Levinson’s claim is that a PI is the considered result of thinking on the part of those who have the power to make such choices.

A performance which instantiates a PI can itself be interpreted, or criticised. Thus there can be CIs of PIs. Propositionally expressed ideas cannot belong to PIs, (a reflectively considered manner of playing), but they can belong to CIs. Thus, an observation about a conductor’s PI is itself a CI:

"Rather it represents a CI of that piece, or of the piece under that PI, or perhaps just a CI of that PI."

These options are significant, because they betray different understandings of the relationship between performance and work. 1) If the critical interpretation of the performance constitutes a CI of the piece, then we are identifying the performance as an instantiation of the piece, as Goodman, and Goehr might wish us to. 2) If we are dealing with a CI of a PI, then no commitment is made concerning this relationship. But the question is begged, ‘what is the PI?’. 3) If we take the view that the performance instantiates a CI of the piece under the conditions of a particular PI; then we are suggesting that the work can have many PIs, which in turn may have different CIs.

1) Under a Goodmanian scheme, a CI of a performance which is perfectly compliant with the score must also be a CI of the work. We should say that it is the CI of the work. If we accept Goehr’s modifications, such that the performance we are judging is not perfectly compliant, but is recognised to be of the work in virtue of the intention of the performers to approach an ideal performance, then the CI can also be of the work. It can also be of the performance, because it is implicit in Goehr’s account that there could be more than one PI which counts as intending to realise the ideal. For Goodman there is only one PI possible, whereas Goehr allows for more than one.

2) Someone might offer a critical evaluation (CI) of a way of performing a work (PI). Here there would be no claim to be judging the work - rather the idea that one was relating a CI of the PI with a CI that one already had in mind. Thus, if one has an ideal performance, one often relates other PIs to it. In this case, the PI to which it is being related (the ideal, or ‘benchmark’ performance) is also being taken to be the ideal instantiation of a CI of the work. In this situation PIs are related to an ideal PI and an ideal CI, where the ideal PI instantiates the CI; or perhaps where the CI is founded upon the experience of the particular, ideal, PI. The PI itself is offered within a tradition and range of other PIs, to compete with them, so to speak.

3) In this case the PI is taken to be the result of considered reflection towards a CI. Thus a performance of The Marriage of Figaro might portray the Countess as the central character of the story, and the PI would result from a considered CI her rôle is actually central. Thus the CI can contribute directly to the way in which the character is portrayed in a given production. This approach acknowledges that there can be other PIs, not only of the same CI, but of different CIs. Thus a PI can represent different CIs; or a CI can have different PIs. Different PIs can represent the same CI; but a PI cannot represent different CIs.

A PI cannot represent two CIs at the same time because any ambiguity in the PI would be resulting from a particular CI which itself advocated a certain ambiguity. Thus a PI of The Turn of the Screw cannot be based upon two CIs, one of which takes the ghosts to be a figment of the Governess’ imagination; the other of which advocates their reality; rather the PI will be based upon a CI that advocates that very ambiguity. For we have said that PIs cannot themselves be ambiguous, whereas CIs can accommodate ambiguity.

Our first case (1) relates to Goodman’s position, or that of any performer looking for the correct, singular interpretation. The second case (2) relates to Goehr’s modification, and allows for comparative judgement among performing interpreters. The third case (3) relates to a more liberal position which recognises a diversity of PIs and CIs for a particular work.

1.4.9 The intentionality of the score

So far we have seen how three of the assumptions I began with may be construed. We might identify the score of the piece with the work of music - Goodman argues this, and although others such as Ingarden disagree, Goehr modifies Goodman’s view such that the relationship between score and performance is intentional after all. We can also see how one might argue that the score is made up of both notated pitches and verbal instructions. We have discussed how artworks are necessarily intentional objects, and how their being so prompts us to enquire after the conditions of their creation, including what the composer thought he was doing, and why.

I distinguished, following from Lyas, the composer in the work, from the composer of the work. When we are involved in criticism, we can succumb to what Beardsley and Wimsatt called the Intentional Fallacy by "consulting the oracle", or we can look for the artist in the work, for whom we need not go outside the work. Now I want to introduce this material into our discussion of scores and the identity of music.

1.4.9.1 Separation of work and score

1) If we decline to identify the score with the work - as Ingarden refuses to - then reference to the score - intentionalistic as it is - involves reference to something other than the work. If we do not identify the score with the work, and suggest as Ingarden does, that they are, and must be kept separate, then there must be distinctions between interpretative judgements of the work and of the score. If we are interested in the music, of which the score is the intentionally constructed symbol system, then reference to the score is not the same as reference to the music. Under the conditions of the Intentional Fallacy, to consult the score is to consult an external feature. The score has the same intentional relation to the work as do documents of a biographical or genetic nature. Under Ingarden’s view, Beardsley and Wimsatt might have to concede that reference to the score was not appropriate to criticism, even though the score is invariably vital to performance.

This view, coherent as it is, seems, like some of Goodman’s proposals, to be counter-intuitive - the suggestion that we should not consider the score to be relevant to criticism of performances is rather strange; and goes against current practice. One of the major ways in which we evaluate a performance concerns the performer’s ‘faithfulness’ to the score; and perhaps more significantly, judgements concerning how the performer has presented ‘the work’. The performer invariably consults the score, so we might wonder why we should not do so ourselves in appraising the performance.

Beardsley and Wimsatt might reply, that even if the performer relied on external information, our task of criticism should not do so. The object of our appreciation is the performance, and we should not introduce what we know about the circumstances of its production, any more than we should introduce what we know of the author of the work. What concerns us, they would say, is the object before our ears.

There is a problem though, caused by the severity of the demand made upon our critical judgement by Beardsley and Wimsatt. They would say that we should ignore what we know about external features of the work. Thus, we should say that we ignore features we know about the composer’s biography; and also, if the score is also external, we should ignore what we know about the score.

Firstly, we cannot ‘unknow’ anything at all, nor can we consciously and deliberately forget it. It is fairly well-known that Berlioz was in love with Harriet Smithson when he wrote the Symphonie Fantastique, and that the ideé fixé was intended to symbolise this infatuation. If we know this, we cannot pretend that we do not. The demand is unworkable.

Secondly, it is more unworkable where our knowledge of the score is concerned. Those of us who know the work may well know the score. It is through the score that we know the work, either because we have studied the score, in association with various performances, or because we have heard performances, which are of the score. It is counter-intuitive to suppose that what we hear is not related to the score, when the score, by any account, is used as a means towards a performative end. If we know the features of the score - the pitch characters and verbal instructions - we cannot pretend we don’t when the aural phenomenon of the performance clearly contradicts the instructions of the score. Thus, if we happen to know that the opening of the Symphonie Fantastique is quiet, inasmuch as the score is marked piano, and we are confronted by a loud noise, we cannot pretend that we do not know that the performers have either ignored what the score stipulates; or have failed in their intention to adhere to it.

The problem here is that Beardsley and Wimsatt’s view that we should be solely interested in the aural phenomenon which we hear, cannot really be allied to a view which separates the score from the work. If they are allied, then we find ourselves being asked not only to forget what we know about the conditions of the work’s creation; but also what we know about the creation of the performance, which includes, under this designation, what we know about the content of the score. This position is unsatisfactory because it is contrary to current practice; and makes an impossible demand upon those who know the score. That such people would probably also want to add that to know the score is to know the work, suggests that we should not fail to identify the score with the work in the first place.

2) If we go with Goodman, and refuse to identify supplementary instructions as part of the notated score-work, then to follow the verbal instructions is to interpret on the basis of external information. This presents us with a very odd situation, in which reference to the notes is acceptable as ‘internal’, whereas reference to the supplementary instructions is not. To criticise a performance for having included wrong notes is acceptable; but criticisms of dynamics, timbre or tempo are not. This delineation, although founded upon a fairly coherent theory of Goodman’s, can hardly be advanced, because it makes a nonsense of the way in which we engage with performances. It also suggests that the only relevant feature of performance is pitch-compliance. This, of course, is Goodman’s view, but we have seen the distinction between the status of the pitch characters and that of the supplementary instructions weakened by Raffman. It might well be true that if Goodman was right, we should have to interpret on the basis of external information. But, we have reason to doubt Goodman here.

1.4.9.2 Identification of the work with the score.

1) If we identify the work with the score, then it must follow that to look at the score is not to go outside the work. Thus neither the treatment of the score as a critical object; nor reference to it when criticising a performance; nor reference to the score as a means to performative criticism constitutes an adoption of the intentional fallacy, that is, reference to external features of the work. Goodman’s idea of chain of performances and scores preserving the identity of the work (which Goehr also adopts), works here.

2) If the score is identified with the work, and the score is constituted by verbal instructions and notated pitches, then reference to the score involves reference to a composer’s expression of intention concerning an ideal performance. In this case, the composer in the work is to be identified with the historical composer of the work, because the score represents the intentions of the composer as fixed at a point in time. The score/work is not only the result of the kind of intention which Beardsley and Wimsatt wanted to remove from criticism; it is also the representative of a specific intention which seeks to enable the production of performance events towards which such criticism may be directed. Hence it stipulates which sounds are to be produced through the manipulation of conventionally agreed upon signs (notation) and verbal instructions. These verbal instructions - intentional as they are - are part of the score, and are part of the work. This being so, we have a situation in which some intentionalistic features are included as ‘internal’ to the work, and I suspect that Beardsley and Wimsatt would have to accept them as such.

The composer cannot be avoided now for two reasons. Firstly, because the work - as instantiated in performances, possesses features which are supervenient upon the character of the artist - those features which Lyas called ‘personal qualities’. Now, to this, we can add that performers in intending to produce a performance of a given work refer to a score, which contains the expressed intentions of the composer concerning how the work should be performed. The performance is deemed to be one in virtue - as Goehr proposes - of the intention to perform the piece, and the recognition by auditors that it does so. The auditors can only decide that it has done so by reference (which may be indirect) to the score. The score in this case is constituted by pitch characters and verbal instructions, which combine (as the score) to make up the expression of the composer’s intentions. Thus the score - reference to which involves reference to the composer’s intentions - is required for the evaluation of a performance. Not only is it needed for the consideration of the extent to which the ideal of pitch-compliance has been achieved - it is also needed where the syntactically undifferentiated verbal instructions are to be considered. Performance involves reference to both; and so does the recognition of the resulting aural experience as being a performance of the given work.

3) Where intentionalistic criticism is concerned, we may wonder what the object of criticism is. For literature we have a text; for painting, a picture. For music, we either have performances; or scores. The objects of musical criticism are usually taken to be - as Goodman might say - performances which are instantiations of the work. Following Goehr, we say that these instantiations are deemed to be such in virtue of their being recognised as intending to achieve an ideal set out by the composer in the score (or through some other means). We do not generally criticise scores - we criticise performances. Under Goodman’s account, there is no difference between these two, in virtue of linkage among scores and performances which preserves the identity of the work through various instantiations. Under Goehr’s modified view, this is still partly true - performances and scores are linked through the intention of the performer to achieve a result that strives for the composer’s ideal, where that ideal is achieved through a performance which perfectly complies with the score.

If the verbal instructions are to be taken as part of the score (and I think they should be, in virtue of the weakness of Goodman’s stipulation that pitch characters be syntactically discreet), then the notion of score-compliance must be extended to include compliance of a performance, not only with pitch characters, but also with these verbal instructions. These verbal instructions are not syntactically differentiated themselves, but I do not see any reason why we should not continue to take Goehr’s line, and suggest that an ideal for perfect compliance be proposed. Once again, such compliance is an ideal, and cannot be achieved, but the performers have an intention to fulfil compliance-conditions, and are recognised to be attempting to do so. Thus performance involves the attempt to comply with pitch characters and verbal instructions. It succeeds if the result is recognised to be such an attempt.

1.4.10 The performer’s duties

Even if the composer’s intentions are manifest in the score, we might well ask why we should strive to give the composer what he or she wants/wanted - i.e. that these kinds of intentions should be honoured. This is to ask questions about the duties of a performer towards a composer. J.O. Urmson has rather helpfully addressed this very question.

He suggests that ethical considerations arise for people contemplating a performance of a piece. The questions which arise vary according to the situation in which they find themselves. He distinguishes between three performance situations, where the performer a) plays the music to himself, alone; b) plays to an audience in a non-commercial context, perhaps to friends; and c) where performers play professionally to a paying audience. Urmson discusses a) and c) in detail, and ultimately suggests that the difference between c) and b) is not very great. He suggests that performers in situation b) have the same obligations as those in c), but that they are "somewhat relaxed." This is so because they do not have the same resources as professionals, or comparable ability. Thus, it is worth saying now that we should consider amateur performers to be governed by the same obligations as professionals, only that they are expected only to adhere to them "as best they can." This is all very well, but it must be acknowledged that a professional may cease to be one; or may choose to play to a group of friends. I suspect that Urmson would want to say that it is the situation which determines the obligation rather than some independent status of the performer. Similarly, then, the performer who plays to himself, may well be a professional practising.

1.4.10.1 Solitary performance

Having reduced Urmson’s tripartite division to two points, we can look at the first - the situation in which someone plays a piece to themselves alone. Such a performer has no duty to an audience, because there isn’t one. I suppose that we should not be distracted by the possibility of someone ‘overhearing’ such a performance, because it is still not a performance to them. Auditors of such an event have merely stumbled across a private activity.

Urmson also believes that the idea of the performer having duties to him or herself is not coherent. A promise made to oneself, or any undertaking, can be annulled by the beneficiary, but if oneself is the beneficiary, there is hardly any sense in the notion of letting oneself out of a contract with oneself. People do talk in this way sometimes; but even such cases have no constraining force.

Urmson then raises the issue as to whether a lone performer has any duty to the composer, and he construes this as being the same question as to whether we have duties towards the dead. At any rate, there might still be duties toward a live composer, even if we had none to a dead composer! Urmson has no difficulty with the idea that one can have duties towards the dead anyway, although he does acknowledge that for some people such a relationship between the living and the dead is anathema. But this is not the issue, he says, rather, the question of duty really pertains to the composer who issues instructions in the form of a score. To what extent must a performer playing to himself, be faithful to these?

To answer this, Urmson appeals to reason:

"The suggestion that the only legitimate use of a score by the solitary performer is to play it as accurately as one can is surely one that no reasonable person would make and one which, I believe, no composer has ever made."

It might be different if a private rendition claimed to be faithful, but there is no insistence that the solitary performer should be required to attempt fidelity. If there were, it would be hard to think of how someone might actually practice a piece, given that there is far more to the practice of a piece than repeated quasi-performances. Such practice invariably involves deliberate deviation from the score.

Urmson therefore believes that the solitary performer has no duties at all as such. He compares the situation to that when one cooks a dish for oneself. One can, privately, call what one has cooked whatever one likes; and can depart from the recipe at will. If one exerts a personal preference, one can hardly be said - in this case - to be failing one’s duties. The distinction between a lack of fidelity to instructions and variation on them is empty under these conditions.

1.4.10.2 Public performance

The conditions of obligation for both professional and amateur public performance do not vary much, as I have already stated. Just as Urmson thinks it would be unreasonable to demand of solitary performers that they have obligations; he believes it is intuitively true that public performers do have constraining duties. These duties can be summed up thus:

"In the end, the performer’s duties to his audience seem not differ widely from those of any purveyor of goods to his customer."

The performer is in a quasi-contractual situation as soon as he or she publishes a programme. He thereby undertakes to play piece X in return for a fee, payable by listeners. Urmson believes that no performer would wish to deny this. But, given that the performer has a duty to perform the advertised programme - how is this to be done? Should he play it in the way he or she would expect the majority of listeners would choose to have it done (the customer is always right)? Or perhaps it should be played in a manner in which the composer would have wished or liked? Or should he or she perform in the way he or she supposes to be the best way? Two problems arise - firstly with the stipulations, which may not be satisfied, since there are often problems with knowing what the audience or composer would like. Secondly, the audience is unlikely to be unanimous anyway - no doubt each position will be represented in the audience. They wish to have the piece performed in a manner that is familiar to them, although that familiarity undoubtedly accommodates a certain degree of variation. As Goehr would put it, the audience would expect to recognise the performance as being of the work. Urmson’s requirement here is rather vague, but properly so.

Hopefully, the performer’s idea of what is aesthetically best will not conflict with this familiarity. This raises the very issue of authentic performance of course, because - as we have seen - it has been suggested by Godlovitch that the manner of performance to which today’s audiences had become accustomed is not aesthetically best. Urmson recognises these issues, but does not discuss them. Rather he says that every performer,

"...must recognise some bounds of authenticity... there are limits, and the performer surely has a duty of honesty to the audience not to overstep them, as he understands them."

Urmson’s basic claim, is that the public performer undertakes a quasi-contractual arrangement with the audience, which he ought not to dishonour. The question concerning the public performer’s relation to the composer still remains, for it seems that the performer "has a duty to the composer not to misrepresent him." But without a standard of authenticity, Urmson believes it is difficult to decide exactly what one’s duties are, either to the audience or to the composer:

"There is no objective, and no intersubjectively agreed standard of authenticity. The performer can only be honest in these matter as he conceives them."

No considerations which we might arrive at, concerning the special relevance of the composer’s intentions, or what sounds best, or what novelties the composer would have liked should be overriding or exclusive. Urmson is perhaps what Godlovitch would call a missionary eclectic, because he deplores performers who evade or ignore these issues:

"I think it deplorable that there should be arrogant performers who believe that the satisfaction of their individualities, their artistic visions, or what you will, are of such central importance that all other considerations can be legitimately ignored by them."

Ultimately, Urmson believes that the issues are too complex for anyone to be dogmatic.

1.4.10.3 Comments on Urmson

Unfortunately, Urmson’s paper touches on more questions than it discusses. His basic view is intuitive - that players performing on their own, in private, can do what they like; while public performers have duties in virtue of the fact that an expectant public has sealed a contractual arrangement by paying money. The interesting question for my account concerns the possible duty that an authentic performer (Urmson’s public performer) has to the composer’s intentions, which, I have suggested, are set out in the score. That is, not the intentions which Beardsley and Wimsatt sought specifically to exclude from critical consideration; but the intentions as to how the score should be rendered - the combination of notation and verbal instructions which themselves make up the score.

But Urmson does not actually argue for any particular duty in this respect - he merely suggests that one might argue that such a loyalty to these intentions is a performer’s duty. In fact he does not advocate any particular duty, his only concern is that no single such duty should become overriding. We might therefore see him as arguing against those who advocate that authentic performance ought to be practised rather than any other mode of performance. It will be remembered that one of the assumptions gleaned from Godlovitch concerned a kind of duty. There was, we saw, a tendency on his part to assume that the composer’s intentions were available through the score, and not so much desirable but required.

Godlovitch believed that the composer’s unique position in relation to his work earned him or her a special status. Effectively, he telescopes considerations about the rendering of pitches into considerations about the rendering of intentions, because he believes that the instructions of the score are intentional. Because we observe pitch stipulations, we should also follow the composer’s preferences expressed verbally. We have seen that this at least, is not unreasonable. We have also seen then, that there are reasons for honouring the composer’s intentions - at least where criticism is concerned. Progressing from Urmson’s account we might suggest that a performer does have a duty to the composer, even though it is one that should not override or exclude other duties. The audience have expectations of works they know, and they expect these expectations to be honoured. If these expectations are honoured, then the details concerning what actually happens are not so important. Thus, we might understand Urmson to be saying that the only duty a performer really has is to face questions of authenticity, or rather, that he or she must be authentic. Goehr might put it differently, and suggest that the performer has a duty to play the score in such a way that the audience recognise the performer’s intention to perform the work through that attempt.

It does not follow then, from any account we have so far encountered, that we must respect the composer’s intentions as manifest on the score. Goodman might tell us that if we do not, we are not performing the work - we are performing something else. To respect the performative intentions can be very helpful - and it makes a great deal of sense if one wants to be recognised as performing the work; but it cannot be insisted that one does so. Similarly, one can gain great critical advantage from considering the circumstances of the work’s creation; the composer’s motivations and why he or she did particular things. But again - as we have seen - it has not been possible to insist that such things be taken into consideration.

Godlovitch’s final assumption, was that we really do have a duty to perform according to the composer’s intentions. If we do so, he proposes, then the instrumentation must be followed contemporaneously, and he then reached the conclusion that authentic performance was desirable in preference to other methods. Urmson and others would want to disagree with the preferment of authentic performance. I now wish to turn to this last assumption.

1.5 The demand for authentic performance

The points I have made so far which relate to the assumptions I attributed to Godlovitch are as follows:

The score and work can be identified with each other.

The score is made up of pitch notations and verbal instructions.

The score is intentionally produced as a means to performance.

The score determines an ‘ideal’ of pitch compliance towards which performers strive.

A performance is identified as being of the work/score if its auditors recognise the intention on the part of the performers to perform that work, and if they accept the degree of pitch compliance reached.

Performers have an obligation to the audience to attempt to perform the work, in virtue of a quasi-contractual relationship between performer and audience.

Performers have an obligation to the composer not to misrepresent him/her in public.

The performer - as interpreter - has a duty to honour the intentionalistic instructional features of the score in order to avoid misrepresenting the composer.

The auditor must refer to the score in order to determine whether such misrepresentation has been avoided.

The auditor - as critic - may also refer to intentionalistic features of the score or performance.

Reference to the composer’s intentions, motivations and background can aid an understanding of what the composer was trying to do by creating a score.

Knowledge about the creation of an artwork is ‘relevant’ in as much as its acquisition satisfies the curiosity we have about the works with which we come into contact.

That curiosity is like the curiosity we have for information about other people.

Some of these points are contentious, but we have seen that they may be argued. Thus, whether they are defended or not, they may be reasonable assumptions on the part of someone such as Godlovitch. That is, they are reasonable, even though I would not want to encourage anyone to assume any of them. If we are being charitable, we may attribute the considered adoption of these beliefs to some musicians.

We have effectively arrived at the position of the missionary eclectic. There is one more assumption which we must discuss, with which the missionary eclectic violently disagrees. The assumption is that the authentic approach should be undertaken in necessary preference to other approaches. The reason, one might suppose, is that the composer knew what he wanted, indicated it to us through the score and elsewhere, and that we ought to do it the way he wanted it done, and, by implication, in no other way.

Urmson touched on this issue in his discussion of the performer’s obligations, but he went no further than to suggest that these matters touch on the nature of authenticity; and that the demand to honour the composer’s intentions is no more than one of many demands which may be considered or adopted.

There appear to be three strands here: 1) We might demand that the composer’s intentions be followed, irrespective of the value of doing so. 2) Or we might demand this in the belief that doing so either a) makes the result valuable; or b) that doing so happens to make it so. 2a argues the necessity of value to intentionalistic performance and carries us into the realm of the dogmatic singularist; 2b renders the value as being contingent. 1a renders the value to be irrelevant.

1.5.1 The relevance of value

Charles Rosen pointed out that authentic performers, in reconstructing the performing conditions of the past in the belief that they are the best conditions for realising the composer’s intentions, are equally committed to bad performances and good ones:

"Like most things, music is generally badly played... I have heard a tape of a new composition in which most of the rhythms were at least slightly wrong... the composer lamented that if this tape were exhumed in the twenty-second century, students would conclude that it represented the performance practice of the twentieth century. As a matter of fact, they would be quite right."

The point is, of course, that we do not by any means perform well most of the time. Any performance practice involves good and bad practice. But as Peter Kivy points out, this establishes performance practice as being value-free, or non-normative. Compositional intentions are normative though, so there is a confusion when we talk about the fulfilment of the composer’s intentions as accepting bad performances:

"For, I take it, we have a right to assume that the composer intends, among other things, the best possible performance of his or her work. Indeed, that is what, presumably, all the specific intentions add up to."

We do not have to be committed to bad performances then - when things went wrong in the past, we can safely assume that the composer did not intend them to. We need not fall into the position of some of Godlovitch’s hodiernists though, and say then that since the composer intended the piece to be played as well as possible, we can legitimately play it in the manner which we believe to be best today.

We do not need to devote much space to point 1a then. Rosen - and anyone else who supposes that authentic performance involves playing badly sometimes - is wrong. That authentic performers do play badly is not in doubt, but it must not be claimed that they should do so because their predecessors did. In the performances of the past - and today - mistakes are made, but in neither case do they fulfil the composer’s intentions. It is therefore a mistaken view to suppose that they do.

The reason for the view may well lie in a confusion between the notion of performances and good performances. A performance should be recognised as being an attempt to fulfil the conditions of the score, and it is true that a ‘bad’ performance may be one that fails in this respect. Logically, it is not bad because it fails in this respect. Following from Goodman, we would rather say that failure to comply to a sufficient degree with the conditions of the score renders the experience a non-performance, rather than a bad performance. This view is taken because it is possible to have performances which do comply, but which we would not wish to call ‘good’. Rosen is talking about failures of notation compliance in his example, but wrong notes or rhythmic laxity are not constitutive value conditions. Rather they contribute to judgements of value; but that is another matter.

It is worth considering whether the composer can intend value at all. The composer may want us to perform well, but, if what we have said about intention is true, then this is not something which he or she has the power to bring about. The same may be true for the features of the score - he may want a performer to play an F major chord, but we may have to deny the composer the ability to intend such a thing. For the composer does not have the power to bring about the completion of such an action - he must rely on others.

It might be said that the issuing of intelligible instructions to those who are known to have the requisite abilities amounts to this. The performer is then the agent of the composer, who may fail his intention if he fails to play the right notes - or fails to play them well. The performer could only fail the composer’s intention to play well if the composer specified how ‘well’ should be construed. This is rarely if ever done, unless one considers the score to be such a specification, in which case we find that our notion of performing the work, and performing it well have been condensed once again.

Kivy may have a problem, because he has introduced a concept of value that goes beyond the notion of notation compliance, but he has combined it with the claim that the composer intends good performances. The way in which Beethoven intends the pianist to play the score of Für Elise is different to the way in which he or she intends Für Elise to be played well. In this position, he may have to acknowledge his own point - that performance and intentional reconstruction are to be distinguished, because the former may be normative, whereas intentions are not. Consequently he may not be able to say that the composer intends the piece to be played well, because this is not something the composer has the power to do.

I think there is a way out for Kivy though. If we say, as I have done, that the composer may be able to intend the notational features of a piece to be performed, and can do so in virtue of the realisation abilities of his or her agents (performers), then he or she may be aware of the consequences of asking them to perform. That is, the composer can reasonably expect the performers to play the notes, because they are able to do so. Urmson suggested that there is a quasi-contract between performers and listeners; perhaps there is also one between performers and composers. The composers expect them to fulfil notation-compliance conditions. When a performer plays in public, he or she is tacitly acknowledging this conventional obligation.

There may be a second expectation, which like the first, becomes a kind of contractual obligation as soon as a performer stands before a paying audience. The composer may reasonably assume that there are performers who are, or will be able, to play the work produced. Not only will they be able to fulfil notation compliance conditions, they will be able to do so in a manner which is deemed by the audience to be valuable. It is not unreasonable to suppose that the composer can intend his piece to be played well, where he or she knows of, or can conceive of, performers who have the abilities (power) to realise this value-laden intention, as his or her agent(s). For example, Brahms wrote his violin concerto for Joachim, a famous virtuoso. The piece is fiendishly difficult in places, but Joachim - and others since - have been able to play the piece, and to play it well. Brahms intended performances to instantiate his work through notational compliance, but he also wanted the performances to possess some value. Brahms himself did not perform the premiére, presumably because he knew that he would not be able to produce a valuable performance himself. Composers often do not perform their own works for this very reason. He could intend a performance to have value only if he could reasonably expect Joachim (and/or later players) to produce performances which would be deemed to be good. When Joachim took the stage he acknowledged this intention of Brahms’. He was in a position of obligation both to Brahms, whom he should not misrepresent; and to the audience, who had paid good money to hear the Violin Concerto in D by Johannes Brahms.

The performer, playing in public, is the agent of the composer, who intends certain things, and with whom there exists a quasi-contractual arrangement whereby both composer and audience expect a recognisable attempt to comply with the notational conditions to be made. If the performer is well-known and has a reputation for excellence, a further expectation of a valuable result may also be present. Such an expectation may be designated through the paying of money by the auditors - the sum determining the speciality of experience expected. Error is harshly judged perhaps, but if the intention to perform the work according to the ideals set out in the score is evident then error is forgivable. Sometimes, such as at a school performance, error will be tolerated to a large degree, and may even be praised.

It is dangerous to confuse questions of intention-fulfilment with questions of value. It is a mistake to suppose that adherence to the composer’s intentions commits us to bad performances, because no composer intends us to perform badly. According to Kivy, he or she intends us to perform well, and we can accept this if we follow Urmson and speak of the performer’s obligations to the composer and to his or her audience. Since the performer is the agent of the composer, there is a sense in which the audience is the composer’s audience, not the performer’s.

1.5.2 Necessary value

The second strand concerns the possible belief that we must perform according to the composer’s intentions (whether or not we do so for the reason I have just mentioned), because the result necessarily has value. Some people believe that the composer was the only person who knew how to perform the work properly, and thus that we can only do so ourselves if we consult his or her intentions. Implicit in these beliefs is another - that there is only one proper way to play the music, and that, necessarily, is to realise the composer’s intentions.

Godlovitch offered as good an account of this part of the assumption as any:

"The requisite devices, the instruments, are taken, ex hypothesi, in the considered choice of the composer as the best tools for the job."

That is, we should follow the composer’s instructions (which include the verbal stipulations of instrumentation) because they are in some sense correct. They may be taken to be correct because the composer says so; or for some other reason. Godlovitch is not clear on this - I suspect that he is not saying that the composer is right because the composer is right - rather that the composer is right because he or she knows what he or she is doing. We should listen to the composer because he or she is an expert in the field of performance of this piece X.

This is a weak expression, and is not as objectionable as a strong view which claims that the composer’s view is automatically correct. However, I have reason to suspect Godlovitch of this view, because - surprising as it may seem, he contends against the missionary eclectics, who, it seems to me, would have no objection to the view that the composer’s intentions are valuable. They might only object to the claim that because the composer’s intentions are valuable, he or she must always be consulted as the oracle. The composer, it may be said, does have a special relationship to the work which yields certain advantages. But Godlovitch seems to want to deny us the freedom to choose to ignore the composer. There is some moralistic overtone in his view, that is too insistent to suggest that he only thinks the composer’s opinions happen to be right.

Behind this debate lies an issue. Does a work have a ‘correct’ interpretation, or can there be many? The missionary eclectic believes that many possibilities exist for valuable interpretation. But others, and it seems that Godlovitch is among them, believe that there is only one way to play piece X, which - even if it is an unachievable ideal - is to be striven at. If approaches are known to thwart the attempt, then they should not be undertaken. No performance has value except one that is ‘correct’, and a correct performance cannot contain mistakes, such as the wrong instrumentation. If one takes issue with the missionary eclectics, it seems to me that one has to adopt this kind of extreme view, even if one is not aware one is doing so. If one argues against a multiplicity of performances (which one may do without advocating an ‘anything goes’ policy, as we have seen from Goehr’s work), then one must be committed to the idea that there is only one way to perform a piece.

Rather than present a defence of a singularist’s position, and argue that multiple interpretation is not appropriate, I shall do the opposite. For assistance here, I shall turn to Michael Krausz.

He says that the differences between what is notated and what is accepted in performance is accounted for by performance practice. Such practices vary through history and:

"...given that there are no univocal or uncontentious overarching standards on virtue of which one may conclusively adjudicate between interpretation practices, it follows that for pertinent cases there can be no singular ideally admissible interpretation."

Furthermore, the incompleteness of the score in relation to a potential multiplicity of interpretations argues against the possibility of there being only one correct one (which one?), and such an ideally admissible interpretation would not be notable (this appears to be an argument against himself!). Krausz believes that ‘extra-score considerations’ are needed in order to make an interpretation.

Even though he does not accept the singularist’s view, Krausz does accommodate singularist views. He suggests, as I have done, that a singularist may defend the position by a reference to the composer’s intentions. He offers a weak example of a composer’s intending something better than what he had written, and suggests that pieces often embody features unintended or anticipated by the composer. Thus he says, the composer should not be given a special status as someone who can determine the meaning of a work, either before or after it has been heard by others.

Krausz’ view is centred on mistakes - his arguments against an intentionalistic defence of singularity is based on the possibility of a composer being overridden by someone else. Often we cannot persuade the composer now, because he or she is dead, but Krausz suggests that a singularist might posit a postulated composer, who might be persuaded that sometime the real composer has done was actually wrong and therefore should be changed. The ideally admissible interpretation could then embody that alteration - quasi-intentional and counterfactual as it would be. Krausz responds to his own objection by saying that if one can postulate a composer who sanctions a change, one can also postulate one who does not. If one postulated composer, why not several? He also wonder’s why we should postulate a person whom we call the composer.

Krausz’ main point though is that since there are a multiplicity of practices, none of which can be established as correct, then there must be several which are correct:

"So, if there is a multiplicity of ideally admissible interpretative practices, it is a mistake to assume that there must be a single ideally admissible interpretation. There is a certain latitude between what is interpreted and what is notated as instruction in a score. The difference between what is interpreted and what is notated is ‘the score in the context of its interpretation practices’.... To insist that the range of ideally admissible interpretations must always be singular is to violate an entrenched feature of classical interpretations, namely the incompleteness of score in relation to the multiplicity of interpretation practices."

This seems to be true, but Krausz hovers too near to the claim that says that because the there are lots of answers to a question there cannot be a single answer. If there are many correct answers, then this would be true; but it would be plainly false if not. That many answers are offered does not make none of them right. Nor does it mean that there isn’t a right answer at all. Krausz cannot mean this; but if he means the first view that there are many answers which are correct, then he has not showed anything, because he has done no more than discuss the assumption with which he began his paper.

Krausz’ paper is not up to our enquiry then - it does not show that there cannot be a singular admissible interpretation, it only says there cannot be. In order to see why this is the case, we need to look elsewhere. For an argument which cuts through the problem I shall turn to Frank Sibley’s "Aesthetic Concepts".

1.5.3 Sibley’s aesthetic concepts

In order to resolve this problem I think we need to turn to Frank Sibley’s "Aesthetic Concepts", not so much because he draws a distinction between aesthetic and non-aesthetic qualities, but because the relationships he establishes are important for an understanding of the view that demands that the composer’s wishes must be followed in performance.

Sibley distinguishes two kinds of remark which we make both about artworks, and things in daily life, claiming that one type can be made by anyone, while the other kind requires taste, or aesthetic discrimination. It is the second kind which concern us here:

"On the other hand we also say that a poem is tightly-knit or deeply moving; that a picture lacks balance... The making of such remarks requires the exercise of taste, perceptiveness, or sensitivity, of aesthetic discrimination or appreciation. Accordingly, when a word or expression is such that taste or perceptiveness is required in order to apply it I shall call it an aesthetic term or expression,"

Sibley is not suggesting that ‘aesthetic appreciation’ is not open to everybody, since: "...almost everybody is able to exercise taste to some degree and in some matters."

The application of aesthetic terms is often supported by other aesthetic terms, but often they are described as dependent upon features which are not dependent upon taste, and which can therefore be detected by any observer with normal sensory faculties. Sibley goes on to say that:

"When no explanation of this latter kind is offered, it is legitimate to ask or search for one... When we cannot ourselves quite say what non-aesthetic features make something delicate, or unbalanced, or powerful or moving, the good critic often puts his finger on something which strikes us as the right explanation."

Thus Sibley claims that aesthetic qualities are dependent upon non-aesthetic qualities. He does not want to say that they are condition-governed though, and contrasts aesthetic terms with words like ‘square’ to show this. Aesthetic qualities have neither necessary nor sufficient conditions to support them:

"There are no sufficient conditions, no non-aesthetic features such that the presence of some set or number of them will beyond question justify or warrant the application of an aesthetic term."

Sibley does admit that there could be negative conditions, since he says that if a picture is painted only with pastel colours it is thereby inappropriate to call it ‘garish’. Aesthetic terms are sometimes characteristically associated with certain non-aesthetic features, but not matter how many examples we come up with, we cannot say that they are governed in any way other than negatively.

When we talk about art, we refer not to general features, but to specific ones, and the reasons we give - the non-aesthetic features which we describe - to support our attribution of aesthetic terms are grounded in particular features of the work:

"... the features which make something delicate or graceful, and so on, are combined in a peculiar and unique way; that the aesthetic quality depends upon exactly this individual or unique combination of just these specific colours and shapes so that even a slight change might make all the difference."

Thus it is Sibley’s claim that aesthetic qualities are dependent upon and are emergent from non-aesthetic qualities, and that the relationship between the two kinds of quality is not conditional (except in the occasional cases where negative condition-government obtains).

1.5.4 An extension of Sibley’s view

What Sibley says about works must also hold for performances of works, if they help to maintain a chain of identity for works, as Goodman suggested. I do not see any reason to suppose that Sibley’s arguments about condition-government do not apply also where interpretation is concerned.

In order to infer from Sibley, we need to posit a three-tier system of evaluation, based on his two-decker construction of aesthetic and non-aesthetic qualities. We need to realise how judgements of value are made, and what funds them. It seems reasonable to propose that judgements of merit or demerit are based upon a consideration of the aesthetic qualities of a work. Non-aesthetic qualities fund judgements about aesthetic qualities which are then taken to be valuable or non-valuable. Thus the blending of violet and yellow on the canvas may result in a grace which the observer finds valuable. If one tampers with the paint, one may affect the painting such that it is no longer graceful; in which case, not only may it not be valued for that quality, it may came to be not valued at all. The same is true where an aesthetic quality - of gaudiness, for instance - spoils a work. If the collision of purple, yellow and pink is altered or removed though, the demerit quality may no longer be present, and the overall value of the painting may be affected such that it is no longer spoilt or overdone.

The same must be true for performances, which as well as attempting to comply with pitch conditions and so on, also possess non-aesthetic qualities, aesthetic qualities, and a degree of value to auditors. The performance, like the work, of which it is an instantiation, is constituted by pitches, which may be rendered in different ways. The combinations of pitches (‘compounded atomic characters’) are non-aesthetic features in Sibley’s view, but the ways in which they are combined yield aesthetic qualities. The melifluousness of a harmonised melody is due to the non-aesthetic combination of notes. The melifluousness may be significantly contributory to a judgement of the work, and of the performance of that work. But the melifluousness, while it is dependent upon and emergent from the notes, is not condition-governed by them. It is possible that certain combinations of notes could not be mellifluous, but we cannot say that certain combinations will always be.

This applies whether we are talking of compounded characters (such as chords) being used together; or of compounds that are whole pieces. We cannot make aesthetic rules for certain combinations of chords, and say that C major preceded by G major will always sound wholesome and complete (even if we think this is generally the case). Similarly, while we may say that Barber’s Adagio is a serene piece, it does not follow that all performances of that work (all playings of those pitch characters under the conditions of the verbal instructions) will automatically be serene, even if they all have been to date. In the aesthetic realm, this kind of entailment does not hold, according to Sibley.

If Sibley is right, and if his account can be moved into the aesthetic realm in general, then we may have a case for saying that condition-government does not apply where doing something in an aesthetic realm is concerned. The value of a performance or work cannot be stipulated either in advance, because there is no logical reason why doing action P should necessarily produce result Q, even though it may do. The reasons which justify our judgements should not be turned into prescriptions, because the logical relationship is not two-directional. We can account for aesthetic qualities by pointing to non-aesthetic features; but we cannot prescribe aesthetic qualities by producing non-aesthetic features.

1.5.5 Contingent value

This suggests that performers cannot demand that the composer’s intentions be followed - they can only recommend it. The view that doing P produces Q where art is involved falls foul of Sibley’s argument; yet that is exactly what has been suggested by those who believe the composer’s intentions must be honoured in performance. It cannot follow from this claim that a valuable result will obtain, if Sibley’s account is relevant here. Doing P (following the composer’s wishes) cannot necessarily produce Q (a valuable performance), even though a good performance may emerge contingently. Authentic intentionalistic performers may recommend the practice, but are unwise to insist upon it, since it appears that what they are claiming is logically impossible. If they relax this claim though, they find themselves at one with the missionary eclectics, who only object to their arrogant insistence that authentic performance is the only means to a correct performance. The view that there can only be one correct performance must be weakened by Sibley’s argument, because the lack of condition-government suggests that one cannot insist on one approach because the results of that approach cannot be known until they are tried. Thus, Krausz is right - that there is a multiplicity of ideally admissible interpretative practices does indicate that there cannot be a single one.

1.6 Summary

We have seen how Godlovitch constructed a defence of authentic performance by positing three points. Authentic performance, he claimed, is what the composer wanted; it suits best what the composer stipulates; and provides the best means by which the score may be performed. He also defended authentic performance practice against its critics. He divided these critics into three groups. Firstly, the Hodiernists, who advocate modern instruments over old ones, and who see themselves as fulfilling the composer’s intentions by playing as well as possible on the most up to date and advanced instruments. Secondly, the Liberal Ecclectics, who argue that what they do is authentic, in that they are merely carrying on the traditions of the past, in the light of the fact that authentic instruments and players are few and far between, and if we prefer modern instruments, then we should use them. Finally, Godlovitch introduces the missionary eclectic, who issues the greatest challenge by advocating both authentic and non-authentic performance as an ideal. Reinterpretation is acceptable if valuable, and so performers and listeners should not be constrained by rules of interpretation.

Godlovitch answers most of these objections, although he is unable to dismiss the Missionary Eclectic. From Godlovitch’s Purist then, we have a challenge - to defend him from having to acquiesce to the tolerant view that different performance practices are all acceptable today. His defence of the Purist is informed by various assumptions though, which demanded to be clarified.

1) Godlovitch assumes that the work of music is to be identified with the score. This view is acceptable, in that it has been held by Nelson Goodman. Lydia Goehr modifies it. Goehr proposes that Goodman’s requirement that performances be deemed ‘of the work’ where perfect pitch compliance obtains, be recast in terms of the pitch compliance being an ideal, towards which performers should be recognisably striving. The view is contentious for two reasons though. Roman Ingarden would disagree with Goodman, since he claims that the score is not to be identified with the work, rather it is an intentionally designated symbol system, not a series of notational characters and verbal instructions. Secondly, Diana Raffman exposes a mistake in Goodman’s construal of notation, such that pitch characters are not semantically differentiated, thus weakening the claim that the verbal instructions are not part of the score, or of the work qua work.

2) The second assumption was that the verbal instructions - which for Godlovitch, include the instrumentation - are not part of the score. Raffman’s point helps us to see that Goodman’s severe, anti-intuitive view should not be adopted, in which case, we should concur with Godlovitch, and suppose that a performance of a score is concerned with both pitch notations and verbal instructions.

3) A central claim of the authentic performers is that they are realising the composer’s intentions. That pieces of music, as artworks, are intentionally created is not in doubt. It is important to investigate the nature of intentions, because some writers, notably Beardsley and Wimsatt, have argued that we should not introduce them to critical enquiry. Their view seems to be weakened by the presence of personal qualities, which are not external to the work, but which involve a knowledge of the composer’s person and intentional behaviour. Furthermore, reference to intentions can be helpful, because what we learn about the circumstances of a work’s creation enable us to form a relationship with artworks - a kind of relationship which is rather like a relationship we might form with a person. In the light of such a relationship, questions of intention take on a certain value.

But our work on intention has been concerned with criticism, and critical activity is to be distinguished from performance. Having separated the two interpretative activities, we can see that it is still best to keep the work identified with the score, and that to do so, enables us to propose that to consider the intentional features of the score is not to go outside the work. This alone might be sufficient to insist upon reference to intentions, but it does raise the issue concerning what duties performers have to the composer, and to the audience. Given a quasi-contractual relationship, suggested by J.O. Urmson, it makes sense to speak of the performer as being obliged to play the work; and not to misrepresent the composer.

Thus there appear to be various good reasons for not dismissing the composer’s intentions, and some arguments to suggest they are helpful. If the performer has a duty to the composer, then it may be very difficult to honour him or her, by ignoring such intentions.

4) The final assumption went on one step from the Missionary eclectic’s position, and assumed that the composer’s intentions must be followed, to the detriment of any other approach. There can be a view that holds that reference to the composer’s intentions necessarily produces the best results. Such a view cannot be held if Frank Sibley is right where aesthetic concepts are concerned. At best, a contingent relation may hold, such that performance according to the composer’s intentions is to be preferred, but the value of doing so is not to be assumed. The last assumption, then, cannot really be held, because it is based upon a logical relation that does not and cannot hold in an aesthetic realm.

We are left then, in the position of the Missionary Eclectic, who has no argument with authenticists, except when they insist on the singular admissibility of their approach. This singular admissibility has been shown to be untenable. Thus, we may say that three of Godlovitch’s assumptions are reasonable - and that in showing that they are, we have seen how the missionary eclectic’s position is a tenable one.

Thus far, we have seen how authentic performance can be defended as being a manner of performance no more or less acceptable than many other approaches. In the second part of my enquiry I wish to consider what authentic performance actually involves, and whether it achieves what it sets out to do. Having defended authentic performance to a large extent, I shall attack it in Part Two.

Part Two

 

Authentic Performance Discredited

 

 

In the Introduction, I described the phenomenon of authentic performance and introduced some of the significant literature on the subject. I made the point that there is little of philosophical merit written, but that most of what there is has the purpose of discrediting the activities of some authentic performers. Having discussed the assumptions upon which authentic performance may be defended, it seems only fair to discuss the criticisms which have been made of the concept. Those who defend the practice are, on one level, seeking justification for what they do. Those who criticise, tend not to want them to refrain from their practices, but prefer to discredit what they take to be the claims of superiority. As we have seen in Part One, there are people who claim that authentic performance is the ‘correct’ way of performatively interpreting a work, but we have also seen that many performers do not claim this much. Their defence is no more than this - they do not seek to attack non-authentic performers. Thus, when we introduce a note of criticism or condemnation of the advocates of authentic performance, there can be confusion as to who is accusing who of what. It is now my purpose to unpack this question.

At this point, it is well to be aware that not only do the defendants take different positions - the attackers do too. The attackers are not always aware of this, and so their criticisms are sometimes not as damaging or as well aimed as they think they are. Sometimes it transpires that no-one actually advocates the practice which is being criticised. Thus one of our problems is to separate a false debate from a real one. In Part One I hope to have shown what the real assumptions and arguments for authentic performance are. In Part Two I shall discuss the arguments employed against the advocates of authenticity; but also indicate the strands of false debate. It will then be my project to engage the two sides in debate.

2.1 Richard Taruskin’s views

In this area we have a wide range of literature, but as with those who would defend the practice, much of that literature is not of specific philosophical value. There is, however, one writer, whose views are coherent, and who should not be ignored. Richard Taruskin has contributed to two scholarly publications, articles which seek to expose much ‘authentic’ practice as little more than a reflection of modern(ist) taste and ideology. The most extensive statement of his views is to be found in the anthology edited by Nicholas Kenyon, but a more brief (and earlier) account may be found in the Early Music symposium of 1984. I shall be discussing Taruskin’s views at some length in what follows, since although he speaks ‘at the top of his voice’, his arguments are certainly coherent and persuasive. The appeal, and the inconvenience, of his publications consists in that he often uses quite strong language, which leaves us in no doubt as to what he thinks, but which also tempts us to accuse him of exaggeration.

2.1.1 Authenticity’ and its alternatives

Characteristically, Taruskin begins his essay in Kenyon’s anthology by saying that the word authenticity is not descriptive nor critical, but ‘propaganda’. It is a word, he says, which has an eternal moral import, and which necessarily privileges one type of performance over another. Thus, the question of preference is ruled out, because ‘inauthenticity’ is a demerit term - to prefer it is to prefer something of less value than authenticity, and to do so indicates some kind of moral defect. Hence, behind what may seem to be an innocent description, or taxonomic criticism, lies a propagandistic attempt to discredit non-authentic performers, who, by simple definition are ‘inauthentic’, or ungenuine. This claim is apparently supported by the inability to define the term. ‘Historically-aware’, ‘historically-informed’, and ‘verisimilar’ all suffer from the same problem when we consider their antonyms, which have negative moral implications.

To these, Taruskin prefers Gary Tomlinson’s value-free concept of ‘contextual’ performance, also advocated by Joseph Kerman. But the chief emphasis is placed on external factors, and risks promulgating the fallacious idea that the recreation of the external conditions of a piece also recreates the composer’s ‘inner experience’ of it. Taruskin rightly cites Christopher Hogwood as someone who seems to believe this about his recording of the ‘Eroica’ Symphony. His recording of Haydn’s The Creation also emphasises an interest in external conditions:

"First recording, with 120 players and 80 singers, to recreate in sound, scale and disposition the performances conducted by Haydn in Vienna. The English text is that authorised by the composer."

Taruskin points out that the attempt to be ‘contextual’ is hindered by the fact that most music is performed today in a concert hall, whereas two hundred years ago performances took place in homes and churches. He would no doubt agree with Nikolaus Harnoncourt, who when interviewed, had this to say about ‘contextual’ performances:

"When I hear the word authenticity it drives me up the wall. Authenticity simply doesn’t exist. I’ve read in several record notes that they are reconstructing exactly the original forces, eight first violins, seven seconds and so on. It is so idiotic. When I read something like that I feel it has absolutely nothing to do with music."

So for Taruskin, ‘contextual’ will not do as an alternative to ‘authentic’. ‘Verisimilar’ is no better for three reasons. Firstly, according to Donald Grout, it belies an interest in what is old and dead, rather than what is alive and new.

Secondly, it remains the case - in spite of recent research - that our ability to measure verisimilitude is speculative, and always will be. We cannot know whether we have achieved it. Various ‘historical’ traditions have evolved recently, but their historicity cannot be verified, and in some cases it can be questioned. The main weakness here is the dependence upon verification:

"Those whose scholarly super-ego insists that everything they do must survive a trial-by-document are doomed to a marginal existence as performers. Strict accountability in fact reduces performance practice to a lottery, for the performer can exercise no control over the state of evidence."

Taruskin compares this approach to religious fundamentalism, whose advocates assume that what is not permitted is prohibited, and which produces results which are specifically intended to prevent the performer from being criticised in the future for having made ‘additions’ to the music. If one wants to avoid being criticised, then one has to resort to doing nothing - and many people seem to be doing exactly that in the name of authenticity. Taruskin both deplores and dismisses such an understanding.

Thirdly, verisimilitude is not desirable because the conditions to be emulated often produced undesirable results in the past. We should not assume that our predecessors performed well, or that we would like what was produced. Peter Phillips has said that the sound of a sixteenth-century choir is hardly desirable today, and he believes that today’s choirs are capable of doing better. The same problem over verification remains though - neither Phillips, nor David Wulstan, who claimed that girls’ voices could produce "exactly the right sound", had ever heard a sixteenth-century choir! Taruskin claims that here is an attempt to pass off a twentieth century phenomenon as ‘historical’. For it seems that the real criteria is one of value - girls sound best but once we have realised this, there is a need to justify the view with an historical explanation. Wulstan - a musicologist - seeks, as Hans Keller would put it, to justify and perpetuate his profession, whereas Phillips is a performer, who, while appreciating the value of musicological research, is primarily interested in producing the best possible performance.

Thus Taruskin puts to one side the terminology of authenticity. He deplores the moralistic language that has come into fashion, although he offers no alternative. Authentic performance that is ‘contextual’, is not strictly so, and he suspects that the value of being so is being extra-musical. Approaches that depend upon historical accuracy, verisimilitude, or historical awareness are desperately hindered by our inability to verify a single conclusion they reach.

2.1.2 Taruskin and intentions

Having cleared the ground of any vestigial meaning of the term ‘authenticity’, Taruskin turns to questions about intention and the responsibilities which the performer has to the composer. Here Taruskin takes the same broad view he has been taking over verisimilitude - the problem, he says, is that we cannot know the composer’s intentions:

"We cannot know intentions, for many reasons - or rather - we cannot know that we know them. Composers do not always express them. If they do express them, they may do so disingenuously. Or they may be honestly mistaken owing to the passage of time or a not necessarily consciously experienced change of taste. If anyone doubts this, let him listen to the five recordings Stravinsky made of Le Sacre du Printemps, and try to decide how the composer intended it to go."

Taruskin’s concern, again, is with verification - what cannot be verified cannot be used in evidence. But this is not his main point - which is that composers do not have intentions of the kind we are thinking of, rather we are dealing with a need to gain the composer’s approval for particular approaches. The appeal to intention avoids the responsibility to understand the music. Just as the quest for the historical music replaces the need to understand with the need to know something concrete, so too does the need to depend on intentions. Both represent what Taruskin terms a ‘failure of nerve’. Taruskin’s concerns are self-avowedly Beardsleian - and he goes on to quote Wanda Landowska’a opinion that once a work has been created, it goes about the world on its own. Even the composer relates to his own work later, as a listener or performer, not as privileged composer. Beardsley and Wimsatt said that the oracle should not be consulted, Taruskin says that there is no oracle. Taruskin considers the matter ‘settled’. But he also considers it to be a red herring anyway, because it is not just authentic performers who advocate adherence to intentions. All performers claim to be faithful to the composer’s intentions.

Taruskin then makes a distinction, between ‘authentic’ and ‘authentistic’, realigning the meaning of the word ‘authentic’ which is generally accepted in discussion. What I have characterised as ‘authentic’ performance, Taruskin labels ‘authentistic’, while preserving ‘authentic’ with a modified meaning. He takes two different attitudes towards the rôle of intentions as being divisive here:

"...the former (authentic) construes intentions ‘internally’, that is, in spiritual, metaphysical or emotional terms, and sees their realization in terms of the ‘effect’ of a performance, while the latter (authentistic) construes intentions in terms of empirically ascertainable - and hence, though tacitly, external - facts, and sees their realization purely in terms of sound."

Authentic performers, according to Taruskin, are interested in reproducing the sounds of the past, whereas authenticists are concerned with conveying the spirit of the piece, not paying so much attention to the means by which the effect deemed to be appropriate is achieved. Authentic performers are interested in the ‘external’ issues - as Kerman’s ‘contextualists’ were, and like Hogwood, aim to get an historically accurate sound. This distinction is characterised as being that between idealism and positivism - the idealist distinguishes between form and content, or spirit and letter; while the positivist maintains that content is a function of form. Both of these positions have histories, but Taruskin believes that the positivists - of whom he takes Wanda Landowska to be a representative - have the upper hand. Some readers may be surprised at this, since some of Landowska’s views seem to suggest - as Howard Meyer Brown puts it, that:

"...she based her decisions about how to play Bach, Handel, Scarlatti or Mozart on her ideas about the essential character of the music, following presumably the spirit rather than the letter of the instructions... Landowska, in short, believed more strongly in her own personal understanding of the music and her commitment to it than in any more dispassionate quest for what the composers would have wanted or expected;"

Taruskin has other evidence though, quoting her as saying that the performer must not change what is written. But while it is difficult to place her - and other performers - on this idealist-positivist spectrum, Taruskin wishes to make the point that there is a misunderstanding between two groups of performers, some of whom wish to honour the intended effect of the music, others the intended sound. At this point in his paper, it his Taruskin’s project to dismiss the account of authenticity which hinges on intentions, but his examples are often wide-ranging and somewhat tangential. He is also wanting to criticise what he calls the positivist approach, which he is quite able to do, having polarised it. The positivist approach (if it exists), is able to judge a performance without hearing it, because it takes the view which we discovered Godlovitch approaching, that not only is authentic performance desirable, it is the only appropriate way to perform older music. Taruskin contests this by pointing out that a judgement which determines the result from a prior knowledge of the instruments involved does not respond to the composer’s tone colours, but to those of the performer’s instrument. This is wrong, for it takes the instrument to be its player, and sometimes goes so far as to equate the (old) instrument with the composer himself. Where this fallacy holds sway, performances are being judged, "not for their accomplishment, but for their class connections."

Thus for Taruskin, fidelity to the composer’s intentions does not produce authenticity. Nor can it reside in the ‘hardware’ - the old instruments themselves - it could only do so if historical verisimilitude could be confirmed. Having expounded these two points, Taruskin leads us to his main point - that students of authenticity who look to verisimilitude or intentions are forgetting that the driving force and main aim of authentic performance is change, strangeness, and the quest for novelty.

2.1.3 The new shock

"And now we have come at last to the nub and essence of authentistic performance, as I see it. It is modern performance... ‘historical’ performance today is not really historical; that a thin veneer of historicism clothes a performance style that is completely of our own time, and is in fact the most modern style around; and that the historical hardware has won its wide acceptance and commercial viability precisely in virtue of its novelty, not its antiquity."

This is Taruskin’s Credo of authenticity - that it has grown up with Modernism and shares its basic precepts. Verisimilitude, fidelity to intentions, old instruments, have been means to an end. They have effectively concealed that end, which has been to provide something new and different. Thus, the shock of the old, is one and the same as the shock of the new where authentic performance is concerned.

Taruskin is not alone in making this observation. In the Early Music discussion Nicholas Temperley said:

"The essential thing here is the belief that what we hear is historically authentic, whether it actually is or not. And if we are to believe that, then the sound must be conspicuously different from what we are used to. Here lies one of the greatest dangers of the movement. For it puts a stronger premium on novelty than on accuracy, and fosters misrepresentation.... many of the performance practices that make up the style have been chosen for their novelty as much as for their historical correctness."

Daniel Leech-Wilkinson hints at this when he asks:

"Is it not far more reasonable to conclude that the remarkable uniformity of approach which dominates early music performance - at least in the English-speaking world - is nothing more than a reflection of current taste - the same taste that infuses present-day performances of Schoenberg, Dallapiccola and Boulez, but which is authentically appropriate only for the last?... It must be apparent, even on grounds of probability alone, that they have as much to do with current taste as with accurate reproduction."

Kerman, in his historical overview of musicology, adopts this kind of view when he says that:

"All this is by way of making the obvious point that the whole issue of historical performance tends to evaporate as we approach the turn of the twentieth century. Historical performance becomes performance... The questions become questions of interpretation. Musicology evaporates into criticism."

For an example of how this is true, we need only turn briefly to Roger Norrington’s recording of Berlioz’ Symphonie Fantastique. The orchestra of Berlioz' day sounds strange to us, and it may be no coincidence that this strangeness adds to the power of the performance. The symphony probably sounded strange to contemporaneous audiences, who, in the fifth movement would not have expected the "deep bells familiar to opera-goers but unknown in the concert hall", but today, the situation is reversed: we know the work, and hence have come to expect (tubular) bells at modern pitch - the surprise comes not because they are present, but because Norrington uses the now unfamiliar sound of ordinary bells. Thus he has given us a new experience of the old.

Drawing on Ezra Pound and T.S Eliot, Taruskin characterises modernism as intransigently advocating precision and the negation of personality - as being "objective, elitist, and fearful of individual freedom of expression." Its goal is to arrest the decay of culture, which can be compared to the (futile) goal of authentistic performance - to "arrest the decay of the music of the past by reversing the changeable vagaries of taste and restoring it to a timeless constancy."

Taruskin goes on to quote Daniel Leech-Wilkinson’s description of how authentic performances display certain common characteristics of dynamics, speed and timbre. Earlier (non-authentistic) performance style is typified by variation in these features, in an attempt to get at what the composer is ‘saying’. Authentic performance is rather characterised by a ‘clean’ sound, and the attempt to avoid playing what is not specifically stipulated in the score (here we may be reminded of the attempt to follow Goodman’s strictures over pitch compliance, and the ‘fundamentalist’ analogy, that what is not permitted, is forbidden). The only justification for deviating from the score (as in cases where there is none), would be to adhere to well-documented performance practice manuals, such as those by Quantz, Leopold Mozart or C.P.E. Bach. As we have already seen, Leech-Wilkinson says that this ‘uniformity of approach’ is nothing more than a reflection of current taste. Taruskin extends this idea in a significant way.

Again quoting Eliot, he reminds us of the way in which the present can influence the past - when a new work is created, something happens to all the works that preceded it. "... for order to persist after the supervention of novelty, the whole existing order must be, if ever so slightly, altered." Thus we can talk of the presence of Eliot in Shakespeare and vice versa, and of Stravinsky in Bach. What is important here is a sense of continuity, in which, as Taruskin’s title puts it, there is a ‘presence of the past and a pastness of the present’. Eliot himself was dealing with the demise of Romanticism, which had fostered a view of art which T.E. Hulme called ‘a state of slush’, by which the beauty of art is equated with its power to evoke a pleasurable empathetic response. Taruskin also associates this view with Eduard Hanslick, Suzanne Langer and Roger Sessions, although he suspects they would probably not admit it, since there were all ‘more or less opposed’ to the idea of music as a language of the emotions.

So Taruskin borrows a concept of ‘vitalist’ performance, typified by Fürtwangler’s and Stokowski’s Bach. The latter employs a harpsichord, but does so with significant rubato. In contrast, there is ‘geometrical’ performance, which tends to abstraction and immutability. Here we have a kind of retreat from the messiness of real things - an attempt to purify through regularity as characterised by a deliberate self-alienation. Taruskin is reminded of Stravinsky, and goes so far as to make that claim that:

"... all truly modern musical performance (and of course that includes the authentistic variety) essentially treats the music performed as if it were composed - or at least performed - by Stravinsky."

There has been a gradual change in performance style which is best understood in terms of a shift from vital to geometrical performance. The vital, or the human, has been pushed out, not by a striving for verisimilitude, historical research or fidelity to intentions, but by a prior demand for ‘objectivity’, precision and straightforwardness. Now we can see the influence of Stravinsky on Bach - from Stravinsky’s geometrical style in the Concerto for Piano and Winds, we cannot help but see Bach’s harpischord concerti at least as potentially disposed towards geometrical performative interpretation. The idea that baroque music is strictly rhythmical has no basis in history - it is an opinion which is "uncorroborated by any contemporary witness". This ‘sewing-machine style, cannot be historically associated with Bach, but it can be with Stravinsky. Taruskin accuses Stravinsky of ‘fundamentalism’ (of which he has already accused the ‘authenticists’), who maintained that:

"execution... (is) the strict putting into effect of an explicit will that contains nothing beyond what it specifically commands."

Again, what is not permitted, is prohibited. Execution is the opposite of interpretation, which Stravinsky deplored. ‘Transmission’ seems to be the purpose of performance, just as Leech-Wilkinson claimed:

"... an attempt to avoid interpretive gestures beyond those notated or documented as part of period performance practice. In a nutshell, the difference is that between performer as ‘interpreter’ and performer as ‘transmitter’."

Nelson Goodman claimed that "complete compliance with the score is the only requirement for a genuine instance of a work," here also Stravinsky seeks an adherence to the letter of the text. Both are from the same basket - a potentially perfect relation could exist here, whereby the composer writes exactly what he intends and the performer does nothing more, nor less, in executing the score. The performer ‘submits’ to the composer, but in such a way that benefits from what Stravinsky calls ‘good breeding’, a sense of tradition which is not merely acquired knowledge. Stravinsky criticises those who would perform the St. Matthew Passion with vast forces (thus would condemn Mendelssohn’s revival presumably), saying that such a disregard for the composer’s wishes betrays a "complete lack of musical education."

Taruskin is interested here in the idea that the first performance has a special status. Stravinsky’s reference to intentions is disingenuous, for he is assuming that the first performance realised the composer’s intentions (which it happens to have done according to Joshua Rifkin), but it need not have done. The point is not that the intentions were realised, but that the first performance "possesses a privileged authority, and that the composer’s wishes are to be gauged in material rather than spiritual terms." Thus Stravinsky is interested in authentic sound not effect, or religious meaning. Remembering T.E. Hulme’s distinction, we can see that what he calls ‘vital’; Stravinsky decries as ‘interpretation’; ‘geometrical’ translates in terms of ‘execution’.

Taruskin’s next step is to introduce the word ‘human’. Ortega y Gasset speaks of ‘dehumanised’ art, and Stravinsky seems to be a proponent of it. Thus we may associate ‘human’ and ‘vital’ if only because Ortega’s description rings true with what we have so far heard. ‘Dehumanised’ art (non-vital art) insists on scrupulous realisation, and ‘iconoclasm’, by which Ortega means the opposite of iconicity. Art itself is of "no transcending consequence". According to Taruskin, this has led to a confusion between the sublime and the beautiful - the beautiful, according to Burke, is "smooth and polished", whereas the sublime is massive, sometimes gloomy. Beauty is based on pleasure; the sublime on pain. Taruskin suggests that with Wagner the emphasis was heavily on the sublime, the great, and that it may not be surprising that current taste has swung away from the pursuit of greatness to the pursuit of accuracy and authenticity. Taruskin himself may be trying to do no more than reverse the swing of the pendulum:

"It became a mission for twentieth-century artists to restore the distinction between bright, wide-awake beauty and blind, irrational subliminity, to reserve the former for art, and to give the latter back to life, nature and religion."

If the non-vital is to be reserved for art, then the idea that anyone is ‘speaking’ in the art is to be rejected. The geometricists deny that it is their voice that an audience hears - they specifically intend not to be heard. Only the music can be heard - the score perhaps - the composer (if he is an ally of Stravinsky) will deny having a voice, and so too will the geometric performer:

"First of all, we try to play the score (of course with all its repeats). Since it is immensely detailed and the work of a genius, it does not seem necessary to add all sorts of extra speed changes or to alter those that Berlioz prescribes... If all these features make perfect sense of the emotional world of the Symphony, so too do the innumerable exact details of note lengths, phrasing and dynamics... We find that the more accurately we reflect such gradations, the better the music sounds."

In his autobiography, Stravinsky said that "Music... is essentially powerless to express anything at all." Hence, "To the proponent of a dehumanized, geometricized art, literally no-one is speaking at all." Authentic performance practice is characterised by this as much as is modern geometrical performance style. The sense of emotional detachment is highly desired by some performers, yet while it may not be difficult to achieve, it is hardly justified on historical grounds. A modern authentic performer may see the work as a construction, whereas a ‘vitalist’ may see it as a statement of faith or a presentation of drama. When an old piece is performed in a geometric way, it is "concomitantly devalued, decanonized, not quite taken seriously." They are removed from their place in the tradition and played in an entirely different way - different to the way in which it had previously been done, for sure; but also fundamentally different to the way in which its creators imagined. Taruskin cannot prove that it is different; but if he is right in equating a twentieth century geometrical style with the performers of both Stravinsky and Bach, then it seems likely - if Eliot was right - that this style cannot be authentic if the word is to be understood as pertaining to past performance practice and/or historical verisimilitude. Thus Taruskin finds it to be an irony that some authentic performers appear to take a moral high ground, accusing their predecessors of irreverence for the canon, when they themselves (Taruskin accuses Hogwood) are no more respectful.

Bach has been "adapted with unprecedented success to suit modern taste". A hundred years ago, Glazunov and Rimsky-Korsakov were bored by the St. John Passion - rather than suppose that our tastes have changed, Taruskin would say that the idea of going ‘back to Bach’ is topsy-turvey. We did not go back to Bach, Bach was brought forward to us. This was no doubt facilitated by the adoption of a universally ‘modern’ performance style, and the massive distribution potential yielded by gramophones and compact discs. What our modern authentic performers have done is no different to what Mendelssohn did with the St. Matthew Passion in 1848 - he reinterpreted it for his own age, and while he may be said to have resurrected it, today’s performers are keeping it alive, for as soon as it becomes too familiar through repeated, geometrical performances, it will become the victim of the form of contempt through which we avoid performances and leave CDs on the shelf.

Taruskin does not see this as a negative observation. Rather he believes that once we have realised that authentic performance is no more than a current performance practice (rather than an old one), we are liberated to see it as modern performance, and thus be appropriately critical. Before we can judge something, we have to know what it is, and Taruskin spends the majority of his words elaborating his thesis that period performance is modern performance. This being so, Hogwood and Fürtwangler, Rattle and Marriner stand on equal footing - none can claim a moral high ground.

2.1.4 The plague of the musicologists

After devoting so much to his main thesis, Taruskin introduces the rôle of musicology, which he sees as mostly negative. While authentic performance was beginning to gain currency, the study of performance practice became interesting in academic circles. The scholarship that arose is described by Taruskin as ‘positivist’. It was interested not in what should be done, but in what was done. And often the ‘is’ was taken to imply an ‘ought’ - an attitude seems to have prevailed such that performers are told to be quiet until a musicologist has found the answer to whichever question they weren’t asking, who can then tell them how to perform a particular (type of) piece. Hard evidence has become necessary, since some musicologists (Taruskin singles out Arthur Mendel) would often attack performers, demanding their historical sources, using the "positive inductive method as a veritable stick to beat modern performers". But this is unacceptable - we should not hold performers to the same kind of account as scholars, mainly because we are ultimately interested in finding out what we like, rather than what it was like.

Hans Keller was particularly scathing about the attempt by musicologists to influence performers:

"What are, superficially, some of the most plausible musicological requirements, making sense to the ear and the brain... turn out, upon inspection and reflection, to be delusions of authenticity such as could grace any ward of a progressive mental hospital welcoming the victims of newly discovered mental diseases, amongst whom the sufferers from musicological delusions would surely figure prominently."

Scholarship can inform performance, of course, and there are plenty of performers who are also scholars, including Taruskin himself. Such people are able to enable us to experience old music newly, but the idea that the mere use of old instruments guarantees a ‘historical’ result is ‘preposterous’. Taruskin wants pluralism, where preference is expressed.

2.1.5 Museums and restoration

Taruskin also wants to express reservations about the idea that old music is to be ‘restored’. We have a view of music that associates it with museums in which performers are curators. Curators deal with artefacts, and while there may be some debate whether the score or music is an artefact, the old instruments certainly are. Just as one can restore an old master, one can restore a harpsichord. But there is:

"... falseness, nay the evil, of the notion, so widespread at the moment, that the activity of our authentistic performer is tantamount to that of a restorer of paintings, who strips away the accumulated dust and grime of centuries to lay bare an original object in all its pristine splendour."

But,

"In musical performance, neither what is removed nor what remains can be said to possess an ontological existence akin to that of dust or picture."

What is stripped and bared are human acts, which exemplify the making of choices, and in such choices consist interpretations. Hence it is interpretation which is stripped away. What people have done - the ways in which human beings have related to the creative enterprise of another human being, is being equated with ‘dirt’. Taruskin therefore takes the idea of ‘restoration’ of pieces to be reducible to the syllogism "People are dirt", and points out that that is dehumanisation.

In conclusion, Taruskin reassures us that not everyone shares this ‘restorative’ view (it remains to be seen if anyone does). Real artists are not interested in elimination of choice and interpretation - they seek improvement and refreshment. At last Taruskin admits what we had begun to suspect - his exposé of authentic performance is not just an altruistic academic favour granted to the musicological community, pointing out that ‘authentic’ equals ‘modern’, he actually admits in conclusion that what he calls ‘authentistic’ performers are not ‘real’ artists (and we can infer that Stravinskyesque musicians - geometricists to a man - are not ‘real’ artists either. Art. for Taruskin, is not about playing the score, or about history - it is about novelty, refreshment and vitality. However, he has criticised his authenticists for concealing those desires. It is not so much that the desires are inappropriate, rather than they are hidden which is a problem for Taruskin, who, to give him credit, praises those authenticists, such as Gustav Leonhardt, who still have the courage to interpret, thereby investing their performances with life.

Taruskin does regard authentistic performances as authentic, because the trends and views he has criticised are not all-pervasive. He is relatively content because he believes that there are people out there who would agree with him. There are people who know what they want to do, and who act on their desires. Such people have values, and are not merely dustmen who take away the rubbish. It is they, for Taruskin, who are the authentic ones, whether they employ old instruments, textbooks and urtexts; or not.

2.1.6 Summary of Taruskin’s views

At this point it is worth reviewing Taruskin’s main points. He believes that most discussion of authenticity arises from false premises. The distinction between ‘modern’ and ‘historical’ or ‘period’ performance is a smoke screen - if one is more modern, it is the historical kind, which is at the cutting edge of modern performance practice. The real distinction is to be drawn between nineteenth-century practice, which he calls ‘vitalist’; and modern/historical practice which is ‘geometrical’. Modern performance is thus modernist performance. The quest for the authentic music is characterised, like its theological precursor, by a quest for novelty. Just as George Tyrell put it:

"The Christ that Harnack sees, looking back through nineteen centuries of Catholic darkness, is only the reflection of a liberal protestant face, seen at the bottom of a deep well"

So too for those who look for historical authenticity in music - they often see nothing more than their own reflection - the ‘reflection of current taste’. That current taste desires novelty predominantly, and authentistic and avant-garde performance style are of a kind, such that the most ardent advocates of High Modernism in music are also the ones who defend authentic performance most zealously. They have become "prevaricators and no longer call authenticity by its right name."

Taruskin challenges the notion of historical verisimilitude - we cannot produce an accurate historical model to emulate, nor can we be sure that we have emulated it. The attempt at verisimilitude is not ‘historical’ at any rate, if historical and modern performance mean the same thing, as he claims they do. Sure enough, there is an attempt at verisimilitude - an attempt to ‘play the score’, but not only can its achievement not be verified, it is not historical. In this much we can imagine Taruskin to be a critic of the Goodmanian view that pitch compliance is to be sought in performance.

Taruskin is an uncritical adherent of Beardsley and Wimsatt and believes that the desire to be faithful to intentions is not a characteristic feature of authentic performance, since "Everyone claims it". The need for authentic performance betrays a lack of self-confidence, or disability to defend one’s practice in any other way.

Nor should musicology be the performer’s dictator. At its worst, musicologists use their research to beat performers with, demanding a positivist’s justification for every artistic move. Taruskin deplores the idea that musicologists should tell performers what to do on the grounds that they have discovered what was done in the past. Here we touch on the idea that modern performance should be a recreation of old performance. Taruskin reserves some of his strongest words for this view, claiming that it is dehumanising to treat a piece of music like an Old Master, with a view to scouring away the ‘dirt’ of human endeavour through the ages. There is a tradition, which itself is part of the history - an approach which tries to remove it is not only anti-human, it is not historical either.

Taruskin concludes with the acknowledgement that those whom he has implicitly attacked are not the sum total of authentic performers. There are some ‘real artists’ who strive for great performances, and who are not bound by a doctrine of performance practice which is most probably modernistic in origin. They know what they want, and are able to act accordingly - it is they who are the real artists.

2.1.7 Comments on Taruskin

Taruskin’s contribution to Kenyon’s anthology is powerful and persuasive. His material is not as clearly ordered as it might be - and is somewhat anecdotal in places, but we are left in little doubt as to what he thinks, and his examples are always relevant. Unfortunately, he neither introduces, nor concludes his paper with any indication of where he is going, or where he has been! He introduces his main theme some fifteen pages in, and his ‘summary’ comes sixteen pages before the end. He offers no definition or description of the phenomenon of authentic performance. His discussion covers some of the main issues - questions about historicity and intention are touched upon; but his main and central argument concerns the ‘newness’ or period performance. This is his main contribution to the debate, and is something for which no performer should presume to ignore him.

But we cannot merely ‘adopt’ Taruskin’s position, for two reasons. Firstly, he is not above criticism himself; and secondly, he has not actually answered the question we are asking here. He has not suggested that authentic performers are doing something ‘wrong’, or that we should perform on old instruments. He might claim that he has knocked down some of the defences authentic performers might put up for their enterprises, but his attacks on ‘intentional’ performance are hardly defended, for instance. Taruskin does not offer us a reason for attacking authenticists - rather he attempts to expose them for who they are. He has made many criticisms and observations - it will be our task to determine which of them are philosophically useful, and which are indefensible, and hence glean material which may be pitted against the perpetrators of period performance.

2.2 Charles Rosen

The American musicologist Charles Rosen took Taruskin to task in his review of Kenyon’s anthology. He meets him in full voice:

"Taruskin writes brilliantly and at the top of his voice, and his most crushing arguments are reserved for opinions that no-one really holds... Still, Taruskin beats his dead horses with infectious enthusiasm, and some of them have occasional twitches of life."

And Rosen is not happy with Taruskin’s ‘narrow’ conception of modernism, which seems to be conditioned by a negative view towards much contemporary art. Taruskin has picked on Stravinsky as an example, but in doing so has ignored the more Romantic tendencies of Berg and Schoenberg which can still be felt today. Rosen seems to be wondering how Taruskin would have argued his case if he had not chosen someone as enigmatic and chameleon-like as Stravinsky as his paradigmatic modernist. Rosen would have a point, but it is worth reminding him that Taruskin also refers to Eliot and Pound, although Rosen’s point is pertinent - the nature of modernism is not as clear-cut as Taruskin would have us believe. In general, Taruskin does not entertain the views of those who do not agree with him, and it could hardly be said that he has considered ‘the other side’ of most of his material.

Rosen also criticises Taruskin for exaggerating the gap between ‘reactionary and progressive’ styles. Taruskin employs T.E. Hulme’s ‘vitalist’ and ‘geometrical’ distinction, but Rosen suspects it is not as simple as that. "They all reach backward and forward". This may be a symptom of a ‘grudge’ - Rosen is surprised that Taruskin will not admit any advances or advantages gained through our greater knowledge of performance practice. If Taruskin believes they are irrelevant, then he certainly does not defend the opinion. Rather he seems to take for granted that such ‘historical’ information can have no value. But that such information is not ‘historical’ does not mean that it is not valuable. Is Taruskin betraying a view that authentic performance is only valuable if it lives up to what it claims to be, but that given that it cannot do so; it therefore cannot be valuable? If so, Rosen certainly wishes to disagree, and even though he spends much of his review agreeing with Taruskin that authentic performance is not quite what it claims to be; he does not say that it therefore has no value.

Rosen’s specific comments against Taruskin are fairly brief; and much of his review article is devoted to his own opinions, which serve as a response to and development on Taruskin’s. It is therefore worth devoting a little space to Rosen’s major points, which may be added to Taruskin’s.

2.2.1 The Rôle of recording

So far, we have not mentioned one of the most significant influences on modern performers - whether they are playing old music or not. Rosen picks up on a passing comment made by Kenyon in his introduction to the anthology, that "the basis of the Early Music movement - aesthetic as well as commercial - is recording." The recordings themselves have the ‘inauthentic’ ability, or inability to do no more or less than sound exactly the same every time they are played. The recording industry has promoted and sustained authentic performance practice, coaxing performers to move from baroque trio sonatas to classical symphonies. Today we have reached Brahms, and even authentic Vaughan-Williams. What Vaughan-Williams would have made of such a performance can only be guessed at, but with some accuracy I suspect. But at least more recent authentic performers have learnt to master the instruments, and so there is no need to record in short bursts so that the orchestras could re-tune frequently. Modern recording technology enables and demands that at least two ‘takes’ be woven together, according to Rosen, ‘objectionable noises’ were edited out. Certainly this would be possible today, using digital processors; but perhaps Rosen is a little cruel in suggesting that retuning was required "every half-minute or so"! His point, of course, is that without such technology, it would not have been possible to produce a half-decent recording of the earlier attempts at period performance, in which case the authenticity bandwagon would never have got going. But since he admits that by now there are many experts, who can play very accurately and attractively, it is probable that the record industry would have picked them up eventually. For the commercial interest in authentic performance is largely confined to the purchasers of that most inauthentic artefact - the compact disc.

The reason for this, Rosen suggests, is that the instruments themselves are not loud enough for modern concert halls. Amplification is possible, but the distinctiveness of the sound is lost. Recording engineers are not faced with a problem here, and balance orchestras and solo instrumental recitals in much the same way. Thus the same power and volume can be achieved for a harpsichord recording as in a Mahler symphony. But in one sense, the ‘different’ period performances have been treated the same as, and made to sound equivalently loud as, modern, larger orchestras. "The more professional an Early Music orchestra becomes, the more it sounds like a conventional one." And today we find the most significant authentic practitioners also working with modern orchestras, such that even the conductors want modern instruments to sound like old ones, and old ones to sound like modern ones.

Rosen accuses the ‘hard line Early music dogmatists’ of making the basically fallacious assumption that a composer only writes (or wrote) for the instruments available to him. Rosen cites an example from the cadenza of Beethoven’s first piano concerto, in which a high F natural is written - which would have been the contemporaneous top note of the instrument’s range. Later in his life Beethoven wrote for pianos with higher registers, and Rosen claims that we know Beethoven would have played an F sharp, because he expressed an intention to revise the earlier composition. But if we ignore the composer’s intentions as Beardsley and Taruskin would have us do, we would either have to play a wrong note in the name of authenticity, or not use the best cadenza which Beethoven wrote for the concerto. As far as Rosen is concerned, authenticity is based, not on common sense, but upon idealism, such that some performers play the ‘wrong’ F natural, even when performing on a modern Steinway piano!

After this digression, Rosen returns to recording - saying that it is no coincidence that recording, which is able to fix the sound of a single performance (although Rosen discounts the fact that rarely is a recording of a single performance-event), has a similar goal to authentic performance - to fix and reproduce sounds of the past. It is not surprising then that the recording industry and the authenticists have paired up so well. Both enterprises reduce music to its sound, abstracting it from its social context or cultural milieu. Behind both is the ‘innocent’ belief that nothing is lost in the process of abstraction.

2.2.2 Original and intended sound

The interest in the sound of the music is new, says Rosen. People have always been interested in how Beethoven played the piano, but not in what it sounded like. Nor did anyone propose using derelict instruments to find out how Beethoven played - instead they read the accounts of Czerny and Schindler. The idea had always been that composition is distinct from the aural realisation of the music. Thus, when Bach or Handel transcribed their own work, they made no attempt to make the new piece sound like the old, but made different pieces out of the same music. The ornamentation may have differed, but this demonstrates the point that ornamentation was considered to be the performer’s job, composition was the composer’s, and so ornamentation was often left entirely to the performer’s discretion (and his or her ability to embellish subtly and ingeniously was a measure of skill and sensibility). Thus Rosen argues that composition and realisation were not identical - "composition was more or less fixed, realization was more or less open." Composition, then, did not involve an attempt to ‘fix’ all aspects of the sound of the piece - the composer did not expect to do so, and rarely attempted to do so. Any composers who did would have been criticised, as Bach was, when he wrote out the ornaments for his singers.

After the Baroque period, things changed, and composers began to stipulate more accurately what they wanted. Berlioz is a good example of a composer who wrote a great deal of detail into his scores. Rosen suspects that the Early Music movement has picked up this need for stipulative detail. But there is a step simultaneously forward and backward involved. The interest is not in what the composer wanted to hear, but in what he did hear. It is progressive, because, realistically, the composer usually writes with available resources in mind, but also has a hope of an ideal performance (in the future perhaps), which "transcends the pitiable means and degenerate practice they have to compromise with." But Rosen claims that many performers have forgotten what happened to music after Bach and Haydn, or after whichever composer whose music they are interested in. Such a disregard betrays a rejection of the modern world, characterised by a rejection of avant-garde music, standard classical performance style, and a break with the common musical life of their fellow performers. Many of these people - according to Rosen -

"... were amateur musicians, for whom Early Music was part of a style of life that included playing the recorder, eating brown rice and wholewheat bread, and making their own clothes."

Rosen admits that now the situation is better - more professional and proficient, but he suspects that there is still something of an inherent prejudice against modernity. This seems like a disagreement with Taruskin, who claimed that authentic performance was about as modern as any performance could be, but both may be reconciled when we consider how a rejection of the modern may itself be typical of current taste. There may be a (subconscious) rejection, as Rosen says; but Taruskin characterises that as being typical of a modernist approach which wants to make everything new. Hence there is a rejection of the new, in favour of the old, when that rejection itself amounts to an affirmation of the novelty implied by the rejuvenation of what was old.

2.2.3 Music and realisation

Rosen argues that music as it once was, should not be separated from the life of which it was a part. Contemporaneous practices involved using a fortepiano as continuo in classical symphonies, and beating time on the ground with a large stick - and both practices have been superseded. In the latter case (and probably in the former) the sound produced was not only distracting, it was not part of the music. Thus, just as Goehr would argue that a ‘violating sneeze’ should not render the performance to be of something other than the work; the beating of a large stick to keep time (which caused Lully to suffer an ignominious demise), should not be considered as part of the music, even if it was part of the aural experience of the performance. Rosen is disagreeing with Goodman here, whose perfect-compliance theory would have to account for such sounds as either part of the music - in which case they are to be included every time; or as extraneous, in which case no performance in which the tempo was indicated would have actually been a performance of the work. This being so, there would be problems for anyone striving to emulate a first performance, because if the work does not include the stick beating, then it did not have a first performance until after the practice was abandoned, in which case the ‘first’ performance may not have been contemporaneous with the composer’s life span. Alternatively, the stick beating should be reinstated in the name of authentic performance - but as Rosen points out, no-one thinks or wants to do that.

Another factor which Rosen considers to be relevant is the difference between pieces intended for public and private performance. This same idea was discussed by Urmson, who pointed us towards a distinction between solitary and public performance. Rosen distinguishes between music played by oneself and friends; and music to be performed for others. We may have the basis of a distinction between playing and performing music here. The ‘player’ does not need to ‘show’ anything by his or her rendition - the major themes or entries in a fugue can be felt in the fingers, rather than heard. If Rosen is right about this, then authenticity cannot claim to be producing what the composer wanted if it advocates the performance of pieces which were only intended to be played. Such pieces were not intended for an audience, and so:

"What the composer wanted or what he expected to get is no guide unless we are content to leave the public in the dark."

In Urmson’s terms, the performer has no obligations to the audience, if he expects there to be none. That being so, the performer should behave as if there is none.

Such ‘private’ music has been performed in public though, and performers have successfully found ways of rendering it in such a way that the entrances of musical themes can be perceived. But this is not a very ‘authentic’ practice, in that its achievement involves ‘stylistic shenanigans’. But to perform the music in public as though one was playing it privately is self-defeating, and psychologically impossible for any musician trained in a conservatory in the last century or so. For Rosen, the ideal way to play Bach’s Well Tempered Clavier is on a harpsichord, by oneself. To ‘perform’ the work is neither practical, nor desirable, according to Rosen. The same problem arises with the ‘semi-private’ music of the Classical period. Music that was written to be played in small gatherings cannot be rendered authentically in a large hall. Rosen gives a detailed and lengthy example of Schubert’s last piano sonata, discussing the conditions of its composition, which need not concern us here. He ultimately concludes that the idea of performing such a piece in public ‘authentically’ today is absurd. Alfred Brendel leaves out repeats, but does so in an attempt to offer the piece today in a form by which it can be appreciated by a large audience. Richter, whom he criticises, performed Schubert’s last sonata (D960) attempting to convey the intimacy inherent in the fact that Schubert wrote it to be played among a small gathering. But:

"Every performance today is a translation; a reconstruction of the original sound is the most misleading translation because it pretends to be the original while the significance of the original sounds have irrevocably changed."

Rosen concludes with another anomaly of authentic performance - today’s audience never behave authentically, even if it could be said that the performers do. In Mozart’s time, it was customary to applaud between movements - perhaps we should reinstate this. Or the convention from Handelian opera whereby people only really paid attention to the arias; went to the opera to be seen; and often chatted through the major part of the evening. It would be absurd to suggest dressing our audiences in period costume, or charging them authentic prices, but not so strange to reduce their number to fit in smaller buildings or to ask them not to sit quietly and attentively and not to be not so ignorant as to clap between the movements of a concerto...

2.2.4 The value of authentic performance

Rosen admits that he considers the ‘basic philosophy’ of authentic performance to be indefensible. He is particularly concerned about the idea that authentic sound should or could be achieved. Authentic sound is insufficient, illusory and undesirable, because sound is dependent on function - the intended sound of a piece (which is often what the composer got, too), depends on what the music was written for. But Rosen concludes on a positive note - believing that our musical life has been enriched by the Early Music movement. It has shown us that there are alternatives to what we have unthinkingly inherited, even if the alternatives which authenticity sometimes offers are not always desirable. We have played Bach, Beethoven, Stravinsky and Dvorak on the same instruments with ‘deadening uniformity’. Taruskin would agree no doubt, although he would want to suggest that authenticity merely proposes a different uniformity, such that we now have a duality where we once had monotony. Rosen is grateful at least for that.

2.3.1 The main criticisms

Between them, Rosen and Taruskin discuss most of the significant critical points about authenticity. Neither of them propose that period performers should stop; or that they are doing something morally wrong. Both suggest that they are misguided. It is difficult to know who ‘they’ are, though, because there is a not a single performer who falls foul of all of Rosen’s and Taruskin’s points. But many of today’s period performers fall down somewhere. Rosen specifically rejects the conclusion that any success that authentic performers have had comes in spite of the philosophy - he says that it is only by attempting to take authenticity seriously that successful, rather than particularly authentic performances are achieved. Having discussed Rosen and Taruskin at length, it is appropriate to briefly review the main points that they make. I shall begin with the points that are more of observations than criticisms.

1 Authentic performance is not historical at all - it is a modern idea which involves the execution of pieces in a modern way. This is not a criticism, but an observation which indicates that a performer who believes that he or she is ‘doing what they did’ stylistically, is misguided. The object which they are doing it do may be identical, but it does not follow, nor does it happen to be true, that what they are doing to the instrument is the same as was done in the past. Not only is this a logical point, Taruskin has demonstrated quite convincingly the similarities between today’s authentic style and conventional approaches to avant-garde music. Taruskin’s view is persuasive, although it is worth remembering that while it may be true that twentieth-century approaches may be consistent with old and new music, by his own admission we cannot know what seventeenth-century performance sounded like, in which case we cannot conclude that because authentic and avant-garde approaches can be characterised equally; a non-twentieth-century approach must have been different. Taruskin does offer anecdotal evidence that earlier approaches were different though, which persuades us to share his view that authentic performance style is more like modern style than the historical practice it seeks to emulate. Hopefully all performers and players will appreciate Taruskin’s acuteness in pointing this out, since an awareness of the phenomenological similarity can inform the practice of the future.

2 The desire to ‘restore’ old music as one might restore a painting betrays a dehumanising influence which smacks of high modernism. Inherent in the hermeneutical attempt to ‘go back to Bach’ and reclaim him in pristine condition, is the assumption that human performers have sullied his work by adding anachronistic performing techniques. These so called improvements cannot be dismissed in this way, because to dismiss the way in which humans have handled the music, is to dismiss the humanity of those doing the handing. It has also been suggested that the Early Music movement involves a rejection of the modern world, such that ‘back to Bach’ becomes an ‘escape to Bach’.

3 Authentic performance is not solely an artistic phenomenon - it has been heavily sponsored as a commercial venture by record companies. Without their support, the current fashion would not have met with the degree of early popularity which it did. Ironically, we notice that the record or CD player is the most inauthentic of instruments, in that it guarantees an exact reproduction each time; yet at the same time it is a most appropriate machine, because it never fails in what some authentic performers appear to want to do - to reproduce a ‘correct’ interpretation, or recreation of a past event. The tape recorder is an ideal tool for someone who wishes to freeze a performative interpretation, such that a mathematical or electrical formula guarantees its being brought back to the life which it once had.

4 The attempt to be ‘contextual’ is doomed to failure. Today we use large concert halls, we have different audience-behaviour, and some ‘contextual’ practices are unnecessary or unpleasant. We do not need to beat time on the floor, or hold the ensemble together with a piano; nor do we want to reproduce conditions of performance which involved amateurs playing wrong notes; instrumentalists’ absence; defective or broken instruments; little rehearsal time; poor leadership; or blatant imbalance or shortage of performers. A great deal of anti-authentic propaganda centres around this point - which is of course valid. It is a little naive though, to suppose that if authentic performers are concerned with reproducing the early conditions of performance, then they should therefore be seeking the bad as well as the good. While critics who make this point are happy to admit that there are hardly any musicians who believe that the recreation of undesirable conditions is worthwhile, these critics still relish their exposure of the absurdity of the desire to recreate whatever conditions held at the time of early performances.

5 It is said that authentic performers seek to perform in such a way that conceals interpretation. It is not possible to perform without interpreting, and so those who claim to do so are on a hiding to nothing. Also, such a performance - if it could be achieved - and those which approach such an ideal would lack value to the same extent that it lacked interpretive interest. Furthermore, an interpretation which seeks to be non-interpretative, is nevertheless an interpretation of sorts. The attempt to avoid interpretation, either by merely playing the notes; or by adhering to the composer’s intentions in a ‘fundamentalist’ way betrays a lack of confidence.

6 The view that authentic performance involves restoration assumes that there is something concrete to be restored - such as a particular and uniquely locatable performing practice for a particular work. It is plainly true that there can be many ways to play a piece - in spite of Goodman’s attempt to deny that many of them would not be of that piece - but we are here dealing with an historical imperialism, but which we have observed, is hardly historical, but typical of a modernist approach that demands execution rather than freedom of interpretation. The charge against some authenticists is that they have no justification for claiming that there is only one way to perform a piece, not only because basic theories of aesthetics show this; but also because our intuition and experience tell us that singularly perfect or correct (modes of) performance cannot be stipulated.

7 Authentic performers are often very interested in producing the same sounds as were heard in past centuries. But the attempt to recreate the sound of past performances is misdirected, in that it ignores the difference between composition and realisation. Composers did not specify performance details, so we cannot treat their music as though they did. A composer would have been aware of the resources available to him; but would also have been aware that a better performance employing better, or different resources could be possible, and so to confine him to what he wrote so rigidly is probably not even respectful of his intentions.

8 The attempt to recreate the sound of past performances is also futile, because one could never know if one had achieved such a goal. Some performers claim that what they do now, or what was done then, sounds better than another way; but there is an insurmountable problem where verifying such assertions is called for.

9 Some music was not written for public performance at all, it was composed for small intimate gatherings, or even to be ‘played’ rather than performed. The idea that such music can be rendered authentically to any audience is anachronistic, and thus the attempt can hardly succeed in ‘recreating’ anything.

10 The significance of the sounds created - even if they could be guaranteed, has changed; nor can it be reproduced. We cannot be seventeenth-century listeners, because we have heard music composed since then. Our listening of Bach is affected by our experiences of Wagner and Stravinsky, and also of the ways in which other people have played Bach. Having heard Glenn Gould, our perception is changed, whether we like his interpretations or not. Other music, and other performances, change us.

11 There is a confusion among authentic performers concerning what the composer wanted, and what he got. Whether we aim to reproduce the sounds, or the cultural context, we need to make a distinction between what the composer wanted to happen; and what did happen. Those who are interested in what did happen, must decide which occasion they are focusing on as an event to emulate. As soon as they begin to collate documentation about different early performances, with a view to producing an amalgamated past event, they have confused this issue, indicating that they are interested in what the composer wanted, rather than what he got; but are willing to accept that what he got is an indication - at least to a degree - of what he wanted. If they are solely interested in what he got - then the problem of verification rears its head. If they are solely interested in what the composer wanted, then it is confusing if they start with what he got; but have to face up to the charge that intentions are unavailable and undesirable. Even if intentions are available through the score, or written documentation, the authenticists should be prepared to accept that since most composers were capable of seeing beyond the contemporaneous limitations of their resources, there is no necessity for authentic instruments.

2.3.2 Recapitulation of the Defence

Having seen the main criticisms, it is now appropriate to remind ourselves of the views on which a defence of authentic performance may be based. Godlovitch claimed to be defending authenticity against its critics, but it is only now that we can really determine whether he has represented his opponents accurately; and also whether they have actually addressed the coherent view that he presents. Rosen accused Taruskin of ‘flogging dead horses’. Having seen both sides of the debate, we can now begin to consider whether the two sides are actually in debate, or whether they are constructing the arguments of their opponents on their behalf so that they can be knocked down easily. For it may be that both, or one of the advocates and critics of authentic performance have caricatured their opponents to a certain extent.

1 Godlovitch defends the musician whom he calls the ‘purist’, and does so by relying on certain answers to particular issues. Thus he seems to take the view that the music is to be identified with the score, in which case pitch compliance should be striven for. This is not as crude as to say that a performer should merely play the notes as written - what is distinctive is that the successful playing of the notes determines whether or not the performance is of that work. The emphasis on the attempt to play the pitches is stronger that it would be for performers who sought to play in the spirit of the work, rather than reproduce it to the letter.

One of the criticisms I have indicated pertains to this point. The idea that it is the performer’s job to play the notes, and not invest any personality, so that the music itself (or the composer through it) may speak, was criticised in criticism 5, above.

2 Godlovitch also believed that the verbal ‘instructions’ written on the score - including the instrumentation - are part of the work, hence that if the composer asks for ‘violin’ or ‘lute’, or ‘flute’, then that is what they wanted, and is what we ought to use, if we are claiming to play that work. If we fail to follow these instructions then we are not complying with the score as indicative of the composer’s intentions.

This view succumbs to criticisms 2, 4, and 11. For here it may be said that we have a hint of the attempt to ‘restore’ old music by using old instruments, and that we should reject modern changes of instrumentation which have developed over the last few centuries. We have also seen that the attempt to be contextual is doomed, which suggests that the value of using old instruments to this end may be questionable. We have also to be wary of the distinction between asking what the composer wanted, and asking what he got, when we are considering the reproduction or recreation of either. We should certainly not assume that they are the same.

3 Godlovitch and many authentic performers believe they have some kind of duty to the composer, to perform the work to which he has put his name. Their performative interpretation should therefore honour at least those intentions of the composer which are expressed in the score. Here is a view which implies that knowledge of what the composer wanted is available to us, and is also normative as a means of deciding how a piece is to be performatively interpreted.

Taruskin and Beardsley would disagree, saying that the intentions are neither available nor desirable, but Godlovitch still has a point if we say that the score represents the composer’s indication of intentions, concerning how the piece should sound. Such intentional information is available, and can be considered with a view to performance. Godlovitch’s purist says it must be.

But this view of intentions must respond to criticisms 4, 7, 8, 9, 10 and 11. As with point 2 the attempt to be contextual is doomed; and the confusion between wanting and getting may need to be cleared up. The attempt to play the piece as the composer wanted it to sound is challenged by the point that the composer may not have had an intention in this respect - he left that to the performer. Some music was not intended to be performed at all.

4 Godlovitch’s purist is of the opinion that the composer’s intentions must be respected, because they necessarily lead to the best performance. The argument is somewhat circular, in that it assumes that the composer has the privileged ability to know what is best for his work; and then concludes from this that following his intentions (wherever possible) must produce the best result.

We have already seen the main problems with this view - it is not philosophically coherent to suppose that there is only one way to perform a work, whether or not we say that that way must involve authentic instruments. We do not need to confront Taruskin and Rosen with this argument, because we have already rejected it, preferring what Godlovitch calls the ‘missionary eclectic’, who values authentic performance, but not so much as to reject other interpretations as incorrect. At this stage it is worth mentioning the extreme purist case though, if only because we have seen that Rosen rejects the notion that there is only one way to play a piece, in criticism 6.

2.4 The debate

Having outlined these apparent conflicts of opinion, I shall now unpack them. After having done so, it will be appropriate to discuss those points (criticisms 1, and 3) which do not correspond to the defence case which I have presented. But first, let us engage the two sides in direct debate.

2.4.1 Non-interpretation

This criticism - that it is not possible to perform without making some implicit interpretive gesture is a strong one, and it seems reasonable to level it at those who attempt not to interpret, but merely to ‘play the notes’. There are few players who merely do this though, although there are many who see it as their job to produce and play from urtexts since they believe that precision of notation and compliance with that notation is the starting place for an interpretation. Thus this criticism is directed at them. It is not that playing the notes is to be deplored - far from it - rather that there is more to playing the music than playing the notes. There is always a danger - sometimes realised - of over-emphasising the importance of getting the notes right, and this goes hand-in-hand with the popular conception, which is also taught to children as they learn instruments, that they should worry about playing the notes, and then about ‘putting the music in’. But the idea that one puts the music into the notes separately is invidious, and this may be what Rosen is getting at when he accuses some authentic performers of trying to go back to the (juvenile) technique that relies only on appropriate technique. It is interesting, of course, to notice that the infancy of the modern authentic movement can be characterised to a certain degree by this tendency, which is most commonly displayed by young performers at an early stage in their musical education. When authentic performance ‘comes of age’, we would expect our novices to infuse a bit of artistic interpretation. And, to give some of our modern authenticists credit - this is exactly what Nikolaus Harnoncourt has done with some of his more recent interpretations of Beethoven’s and Schubert’s symphonies.

Furthermore, Leech-Wilkinson and Taruskin have suggested that this style of (non)-interpretation is actually typical of modernist ‘sewing machine’, geometrical technique. So not only were the young advocates of authenticity playing as children learn, so too were their elders and betters, for whom the avant-garde may well have amounted to the onset of a musical senility, typified by an equally juvenile approach. With them, however, we have no debate. But to the geometric authentic performers, we may say that their apparent lack of interpretive style is not only impossible to achieve; but that actually it is not hard to characterise it in terms of an easily understood approach. No-approach is itself an approach, and is one with which we are not only all familiar, but of which we have been advocates, however unwillingly, in our youth.

The Godlovitchians would want to defend themselves, of course. They might, having accepted that the Goodmanian concept of complete pitch-compliance is not workable, adopt Goehr’s view, and propose that the recognised attempt to comply with the score constitutes a performance of it. Thus they might say that they seek to render the notes of the score to the best of their ability, aware that subtle changes (either deliberate or intentional) do not make any difference, for as long as the audience accept their offering of performance X as work X. To admit this though (which seems to be the only coherent thing to do), is to admit that a degree of interpretation is possible, and certainly not forbidden for as long as it does not render performance X unrecognisable. This then lets them off the charge of being ‘fundamentalists’, who only permit what is stated, and who reject areas of doubt or ambiguity. Goehr’s view admits ambiguity though, and so if our performers accept her point of view, they are not so harshly judged by criticism 5, because they allow that there can be more than one performance Xn which counts as a performance of piece X.

This has implications for criticism 6 - that it is unreasonable to propose that there can only be one interpretation for piece X. If one accepts Goehr’s view, then one cannot make this proposal. The only way would be to take a view similar to Goodman’s, which we have seen is unworkable and counter-intuitive. That Goodman admits that his view is unworkable and counter-intuitive does not absolve his theory, it rather supports his opponents, most of whom are directly concerned with the very practical business of performing music. A theory which says that it cannot be done (inasmuch as perfect pitch compliance cannot be achieved), is not very helpful to any performer. This being so, it seems rather odd that critics such as Taruskin presume that there are performers who advocate the kind of theory which permits only one performance X to be of piece X, and which renders the vast majority of their attempts futile, enslaving them to a pointless and probably valueless idealism.

For there is the main question - what is the point of such an interpretation even if it could be achieved? If it lies in geometric technique, with the music taken out, so to speak, what value does it have? If music is more than just notes, then logically, it cannot merely have a musical value. That need not be a problem, but it means that our authentic performers would have a very different kind of job than their colleagues - unless, of course, their colleagues are all called Stravinsky. If, for example, Christopher Hogwood wants to perform non-interpretatively (let’s suppose he can) - he might produce something of aural interest, but it would not be of aural interest because it resembles aural experiences of the past - which it would not - at best it would be of interest because it was new and different. Actually I think it would have a different interest again, for given that a non-interpretive performance is impossible, the performance would have value only because it had achieved something that had not, and could not be done before.

But here we are in the field of straw men - for I doubt that any performer wants to eliminate interpretation - even if some critics rather like the idea that they do. Some performers seek to reduce the degree of interpretation, but this idea betrays a notion that concurs with my opening point - that technique and musicality can be separated - even worked on independently, and this is hardly something that composers and performers of the past would have accepted.

So, to conclude on this point, it seems likely that most performers tacitly accept Goehr’s view on pitch-compliance, which accommodates ambiguity and a range of performances, which means that any who advocate a single best interpretation (such as Godlovitch’s purist), are probably contradicting themselves, but also that the criticism is rather hard on those who do not. The majority of authentic performers do not try to non-interpret, but the message we might have for them is that they should be wary of attempting to separate technique and pitch-compliance from more artistic intentions which any composer before Stravinsky would expect them to exercise. Ultimately though, the charge we have indicated in criticism 5 is an exaggeration of what is attempted. The main concern is of the value of these attempts - the attempt to be non-interpretive may be of little musical value, but for those who accept Goehr’s view, we find that the value of the result is proportional to the extent to which the range of interpretive possibilities is engaged with. As Taruskin put it - the valuable authentic performances tend to be produced by those performers who know what they want and what they do not want, and act upon it. When this happens, the decision about value is handed to those who listen to the result, rather than predetermined by the interpretive stance of the performers.

2.4.2.1 Restoration and instrumentation

Some authenticists are accused of attempting to ‘restore’ old music by using old instruments, and of proposing that we should reject modern changes of instrumentation which have developed over the last few centuries. One of the reasons for criticising this idea is that it is perniciously modernistic, seeking as it does to remove the influence of centuries of adaptation and interpretation. This ‘removal of dirt’ was described by Taruskin as being dehumanising. Are authentic performers really seeking to dehumanise their art?

It seems reasonable to suppose that most authentic performers would immediately wish to deny the charge, since it has a negative moral implication. The charge is much stronger than any which says that authenticity represents a quest for novelty, since there is nothing of which to be ashamed in searching for the new. But to dehumanise art is presented by Taruskin as some kind of crime. But the question is not so much whether to dehumanise art is criminal - but whether the accusation is actually true. Taruskin’s view is certainly a clever piece of argument, but it also smacks of propaganda and exaggeration. For it is certain that even if he is right, there is no intention to dehumanise in which case we should read him as issuing a warning, rather than describing the authenticists’ project.

If there really is a naive assumption that by restoring the instruments - the objects - the music is itself restored, then the authenticists stand rebuked. The music is not an object which can be restored, like an Old Master. But who has claimed that it is? If they have, then Taruskin defeats them, but this charge does not engage with the many authentic performers who make no such analogy. They see the restoration and copying of old instruments as a means of recreating the sound; of offering something different; or even of making an historical performance. These three aims have objections - as we have seen, but none entail the idea of restoring an object, even though they may involve this idea. The point is, that Taruskin’s observation does not deal with the majority of performers who believe that if the composer wrote ‘flute’, then that is what he wanted, and can and ought to be provided. There is no necessity that such performers must also see the work as an old master to have dirt scraped off.

Taruskin is only right if he is dealing with the purist who advocates a single correct performative interpretation. Such a purist might well say that old instruments should be used because that is what the composer wanted, and that what the composer wanted is paramount, and that the only way to realise that is to give the work the instrumentation specified, and to read instrumental instruction contextually. Such a person might indeed be wanting to reject the performance practices of later years, suggesting that what later generations did to the works was detrimental. But we are no longer debating with the purist, but are rather debating with the missionary eclectic, who would not want to argue for a singular interpretation, and who would admit the potential value of diversity. So Taruskin’s argument is with someone whose position we have already seen to be untenable.

The purist might have a reply for Taruskin however. The missionary eclectic might also wish to join him in accusing Taruskin of holding a dehumanising view himself. For we have seen that Taruskin adopts Beardsley and Wimsatt’s view that intentions are not to be considered in interpretation, because they are unavailable and undesirable. But beneath this view is implied the idea that the work of art - the human creation - severs contact with its creator, at birth. The human involvement is lost immediately, and the artwork becomes an object, rather than a human object. Yet, as an intentional object, it retains a vestige of human content - Beardsley, Wimsatt and Taruskin seek to ignore this. The human features are inconvenient and unavailable.

If what we have already seen in Part One is correct though, then not only is this idea circumvented by personal qualities; we also have a tendency to treat artworks as though they were human, and this suggests that it is not the performers who seek to respect the composer’s intentions who are dehumanising the work; but actually Taruskin in not allowing this human element to be considered. Taruskin himself could be accused of taking a dehumanising attitude - not which removes the dirt of later interpreters; but which seeks to remove the human ‘dirt’ of the composer himself. In treating the work as independent of the human creator, the human is removed and this is no less dehumanising than the attempt to remove later human influence on the art.

Art is a human enterprise - Taruskin rightly points out that it is absurd to ignore this, or to seek to remove the human. But at the same time he adopts a view of intention which tacitly contradicts his desire to reinstate the human interaction with the work. If the performers are allowed to be human in interpreting; then it seems reasonable to allow for the fact that the work they are engaging with is the product of human activity, thus that when the composer stipulates something he does so for a reason - and that even if we cannot discover what he had in mind, we can at least allow for the fact that he had something in mind.

On this question Taruskin does not win. His criticism is directed at only a small proportion of performers, whose position we have already called into question for much more significant reasons. His charge does not touch the others, who might want to return his allegation, suggesting that inasmuch as he believes that the intentions of the artist are unavailable and undesirable, Taruskin himself is denying the human origin of the work, even though he seems to want to affirm the human response to it.

2.4.2.2 Context and verbal instructions

The defence we are considering posits that the verbal instructions on the score represent the composer’s intentions concerning how he would have liked the music to be played, and that where instrumentation is specified, we should read it contextually such that ‘flute’ on a Bach score means ‘eighteenth-century flute like the one Bach would have had available to him’. Many authentic performers would not necessarily demand that a baroque flute be used - they would recognise the modern alternative, but since they are striving for the particular kind of performance which they call ‘authentic’, a baroque flute is desirable in order to preserve the integrity of their approach. ‘Authentic’ here involves honouring the composer’s intentions, especially with respect to instrumentation and other features which can be reproduced.

The objection to this enterprise is simply that the context cannot be reproduced, and that even if it could, it would not be desirable because negative factors would also have to be recreated. Thirdly, there are some contextual features which are no longer necessary.

What is unclear in this part of the debate is why the honouring of the composer’s intentions should extend so far as to include factors which he or she would not have even considered stipulating. Jean-Baptiste Lully, for instance, would not have made any stipulation about beating time on the floor with a big stick, because it would not have occurred to him to do so. Perhaps he would assumed it would happen anyway - but this does not mean that he wanted it to happen. This touches on criticism 11, of course, because we may be confusing what the composer wanted with what he got.

We could clarify the question by saying that we are actually only interested in honouring expressed intentions (taking instrumentation to be an example). If we did, then the issue of time-beating is outside the realm of interest, in which case we do not do it (as happens with most authentic performance). But if we do that, we must concede that our authenticity is not comprehensive, and is limited to the areas which we feel are important. This being so, we would hardly be claiming to be contextual.

If we presume to judge whether or not the composer wanted time to be beaten out on the floor, then we can look at documents, and consider what the composer got; and we may be brave enough to decide that the composer got that and wanted it. This would be saying that we have every reason to believe that the composer was happy with contemporaneous performance practice. We might be braver still, and dare to suppose that the composer was happy for time to be beaten out, because it was necessary to hold the ensemble together, but that he would have been delighted to find a group of musicians who did not need to be led in this way. Thus we might say that we were honouring the composer’s intentions by not conforming to contextual practice. This is, of course., what happens today; but it is ironic that even though a conscious decision has been taken not to thump on the ground, the claim both of respect for intentions and contextuality may be made. Yet it seems that both claims cannot be made. If one honours intentions, thumping is to be dispensed with; if one takes context as ruler, thumping really should be included, no matter how distracting is the result. So it is not clear that honouring the composer’s intentions and being contextual are compatible.

If it is the case that one can only be one or the other, we are left with the desire to be contextual, independent of the composer’s intentions. Christopher Hogwood seems to be a good example of someone seeking this - for him context is the key, and information on it is to be gleaned from information concerning what the composer got - not from what he wanted. For Hogwood, there is value in recreating the context, and in doing so in a way which he would also describe as being vital (but not necessarily in Hulme and Taruskin’s sense of the word). What the composer got is linked to what he asked for on the score - not because it is what he wanted, but because by convention performers played on what was stipulated. When what was stipulated was not available - as in the case of the first performance of Beethoven’s 3rd Symphony, Hogwood famously copies what Beethoven got, not what he wanted.

Against this enterprise we find the question being asked - why do this? The missionary eclectic performs contextually - without insisting that anyone else do it - but what is the value of doing it? There are two answers he or she might give. Firstly, that it is different, and has value as being a new, or unusual way of playing familiar music. The second answer might be that it has historical or pedagogical value - that doing so shows us how it was then. There is a weak and a strong form of this answer - the stronger form insists that it is possible to do this, and that the old experience can be recreated given knowledge of the early conditions of the first performances of a given work (this is probably very close to Hogwood’s view). The weak form acknowledges the difficulties with the attempt to be contextual, but still defends it on the grounds that an approximation of the past is better than nothing.

Concerning the first answer - here we have an agreement with Taruskin’s main point. Where the performer and our critic agree, we hardly have a debate, and there are indeed performers who have accepted the point that authenticity represents novelty, and who are happy to proceed in this knowledge. Many missionary eclectics would take this view, suggesting that it is interesting and valuable to study the historical context and to recreate it as much as possible. Thus, having given the first answer, they might proceed the add the second answer - in its weaker form - acknowledging the novelty of their approach, but also arguing that it has some value other than mere difference. The difference, they might say, resides in the ability of authentic performances to show us at least something of how it was; and also gives us a wider range of instruments the use of which to learn from and experiment with.

The weak form of the second answer acknowledges that the attempt is futile, but still advocates that there is value in trying. Against this we have little complaint - for if the perpetrator is aware of what he or she is doing, then we have a valid and thought-out activity, which is aware of the limitations imposed upon it by the exigencies of modern life. Such performance can give us some idea about how music was performed in the past, and can give some idea what it sounded like. As long as the claims made for such an enterprise are not too sweeping, such an approach can be accepted for what it modestly tries to do. In trying, the realm of our knowledge of the music we love is increased. The interest in historical conditions is acceptable to anyone who takes the view that we relate to works of art in a similar way to the way in which we relate to people, thus that our desire to know about the work - because we love or like it, is natural, and that if we find such performances interesting or valuable, then they are acceptable as alternatives in a range of possibilities. The advocate of the weaker attempt at contextualisation will not disagree, and so here we have no conflict at all.

Concerning the strong from of the second answer, it is argued that the attempt to be contextual is futile because it cannot be completely achieved. It cannot be achieved because we cannot know what all the performing conditions were. If guess, then we cannot be sure we have succeeded. Our ability to verify verisimilitude is speculative, and so cannot be defended. In reply though, it could be said that not knowing we had succeeded does not mean that we therefore cannot succeed. We might succeed without knowing it. But the objector replies that there is hardly value in attempting something the success of which cannot be measured.

The stronger arguer might then say that such performances can be successful and valuable to their listeners. The value though, does not reside in the successful achievement of authentic context - the value lies in the novelty; the musical value; or in the recognition of the attempt, which would also be present for a listener who listens to a performance put on by someone who acknowledged that complete contextuality could not be achieved. The performers may gain the false impression that their performance is valuable because it demonstrates the context; but that it is popular rather in virtue of some other value it possesses. Performers are not always the best judges of why their performances are valued, because they may fail to achieve what they set out to do, and achieve something else, and still be unaware of it.

In this argument we find once again that the purist, extreme view can be criticised, because it advocates the impossible. But the view which accepts the limitations of its approach cannot suffer when those limitations are merely restated by its opponents. The missionary eclectics do not suffer if they are prepared to admit that complete contextual fidelity is not achievable, and cannot be criticised in return if they also accept that what they are offering represents no more than an attempt to educate and discover something of the historical context, when what is gained is valuable to those whose relationship with the music consists partly in wanting to learn as much about it as possible. Ultimately then, our missionary eclectic emerges relatively unscathed from this charge; whereas the idealistic purist would have to account for the futility of his or her activity.

2.4.2.3 Composer’s intentions and the score

The third charge which pertains to the view that instrumentation is part of the verbal instructions which are part of the score concerns the confusion between what the composer wanted, and what he got. We have touched on this already in the previous two sections. If the score does represent the composer’s intentions concerning what instruments he wanted to be used (and we have seen that this is a coherent view), then we must deal with the problem of how to realise those intentions. For the challenge from opponents is that the main way by which these intentions are realised today comes through the study of historical performance practice, in books mostly, and that this enquiry indicates not what the composer wanted, but what he is most likely to have received from his performers. It is not appropriate to suppose that the composer was always happy with what he got, and it seems reasonable to suppose that at least in some cases there were composers who would have wanted different instrumentation that what they were able to specify. If the defence consists in the claim that the composer was the best judge of what was best, and that what he got indicates what he wanted, then we need only find one case where the composer was consciously looking forward to ‘better’ instruments, and the defence falls down. Such an example has already been taken from Beethoven’s piano sonatas, some of which contain dynamic markings which even today are unplayable.

But we should be wary of caricaturing the players of authentic instruments. For it is quite possible to study what the composer got, in the context of an enquiry into what he wanted, without thereby supposing that the two are the same. It is of course, also possible to confuse the two, and inasmuch as anyone does confuse the two, the criticism is valid. We do not need to take the opposite extreme in order to avoid the mistake, and suppose that information concerning how music was performed is irrelevant to an attempt to do justice to the composer’s instrumental instructions. Intuitively we surmise that if the composer asked for a lute - he intended his piece to be played on one. It does not follow that he would not have liked it to be played on anything else. If we do not know what a lute is, or how to play one, then it makes sense to refer to contemporaneous manuals or reports to find out, and in doing so, we need not be confusing wanting and getting.

Once again we find that the criticisms are valid, because it is possible to make the mistake of holding the extreme view which they attack. But it is also possible to respond, either by taking a less extreme view; or by expressing an awareness of the pitfalls. That one may confuse the issue does not mean that one will, and perhaps the missionary eclectics will be grateful for having this danger pointed out to them. But they can continue to use old instruments in the knowledge that whether or not the composer would prefer them today, they are at least approximating the conditions of the past, in such a way that may have a degree of historical interest and value to music lovers. Generally though, we find that the criticism is pitched at the extreme case, and leaves many modern authenticists untouched.

2.4.3.1 Context and intention

Here we confront the view which has so far been implied - that the intentions of the composer are both valuable and available to someone wanting to make an interpretation. One of the charges against authenticists who believe this, has been well-rehearsed here and elsewhere - which is that such intentions are not available or desirable. Another is what we have just seen - that we should not confuse what the composer wanted with what he got, and we have heeded this and realised that authentic performance is still possible in the light of this warning.

Another popular charge - which we have not so far discussed in detail, is the charge that adherence to intentions involves adherence to defective or badly judged intentions. Thus a composer may write something which is not best realised in the way in which he has set it out. Hence the commitment to intention also involves a commitment to the composer’s failures. Implied in this criticism is the view that a piece should be performed in the way that sounds best, not necessarily the way in which the composer wanted it done. Sometimes the composer is not the best judge, in which case we should decide which style of performance suits our age best. This is a plea for freedom of interpretation, emerging from a view which seems to suppose that fidelity to the composer’s intentions can only produce one interpretation, and that if that interpretation is authentic, it must accommodate negative and positive aspects.

But this is surely a misconception - that if one is faithful to the composer’s intentions, then only one result can arise. This would only be true if the composer’s intentions were so comprehensively available that every detail could be specified to the performer. And we know that this could never be. Fidelity to the composer’s intentions involves the consideration of the intentions which are available - through the marked score; through external documents; and through personal qualities. No-one can demand, or claim that they are being faithful to intentions which have either been lost, or which were never expressed in the first place. If the amount of intentional information is not comprehensive (and it never is), then one can be faithful to those intentions and yet still exercise a certain degree (sometimes a large degree) of interpretive choice. Only the radical purist could advocate complete compliance to intentional data; but we have seen that even so, there would be areas of interpretation which would require some decision to be taken, in which case the ideal is not achieved. So we cannot suppose that adherence to intentions involves some kind of restriction, inhibiting the free activity of the performer.

If criticism 7 is to be heeded, then we can see that no composer would ever have intended his intentions to be adhered to so strictly. For we have learned from Rosen that the distinction between composition and realisation were conceived of separately, and that composers often expected certain features of performance to be dealt with by the performers, and that the way in which they did so was a mark of their skill and sensibility. So no composer would have intended his intentions to be adhered to so faithfully as some modern performers suppose.

So here we find that our criticism that adherence to intentions involves us in following the composer in his bad judgement, falls flat, because there is no reason at all why adherence to intention and singularity of interpretation should be associated. It is often within the range of what we know about compositional practice to take the view that a composer would not expect everything he stipulates to be adhered to, and that there were many unstipulated features the execution of which was left entirely to the performer’s discretion.

There is also a misconception that fidelity to the composer’s intentions must involve what Taruskin called fundamentalism, such that the intention-faithful performer cannot make a decision on which the composer could not have expressed an opinion. Sure enough, we have seen those non-interpreters who dare not do anything that is not stipulated in the score - but we have already seen that they are mistaken if they believe they can perform without interpreting at all, also that even if there was a composer who presumed to stipulate every detail, he could not succeed, which would mean that some interpretation was required. Berlioz, it may be said, asked for less - giving the performers more to read and less to worry about - but it is not true that he wrote everything into his score, so that all the performers had to do was play what was written. Issues of tone, tuning and any issue which deals with the acoustics of the building were not and could not be indicated by Berlioz. In performing the Requiem or Te Deum factors such as the reverberation in the concert hall or cathedral need to be taken into account, and Berlioz himself could not have indicated these, not least because some of the buildings in which these works are performed today, had not been constructed in the 1830s.

The criticism that adherence to intention involves adherence to bad intention, or the deliberate ignorance of issues on which no intentions were expressed is unfounded, in that it deals with a narrowly selected idea of what the composer’s intentions were or could be. Neither does it allow for the fact that the composer might also intend us not to follow what we take to be his intentions too rigidly, he might even have specifically wanted us to exercise some judgement in those areas over which he had no desire to exert any authority.

2.4.3.2 Intended sound

Here we find the assumption that the composer intended the work to sound a particular way up against criticisms which deny that we could or should attempt to recreate the sounds of the past. 1) The attempt is misdirected, because it confuses the distinction between composition and performance. 2) It also fails to take account of the fact that some music was not written to be listened to at all. 3) Thirdly, the attempt to recreate authentic sound is pointless, because we could never verify the achievement. 4) Finally, we should not forget that the significance of those sounds then and now is different, hence that even if authentic sound could be achieved, it would not produce the same effect in listeners.

Point 1) has been addressed in the previous section, when we saw that some composers would have left the business of performance to the performers, whose job was seen as distinct. Thus it may be appropriate to talk about intended sound, when the composer had no such intention. But of course, this does not hold for all cases - there were and are still plenty of composers who did have a particular sound in mind, and who notated the score accordingly. This will bring us to points 3) and 4).

Point 2) can be discussed briefly, since it is related to point 1). Again, it may be said that some music was written to be played by a musician alone - at best in the midst of an intimate gathering. Such cases are exceptions though, and do not hinder the basic question concerning whether or not it is desirable or possible to recreate the sounds of the past. Even if one admits that there are cases where one should not attempt to do so; then it becomes appropriate to accept that authentic performances of the Well Tempered Clavier are anachronistic. This seems to be a fair point, but it ignores the fact that whether anyone listened or not, those early ‘performances’ or ‘playings’, still sounded like something. It may be anachronistic to perform them now, but if one is genuinely interested to know what they sounded like in that context, then the criticism that the composer would never have wanted anyone to listen hardly prevents modern attempts to recreate the intimate experience for larger numbers of people. We do not know that the composer didn’t intend an audience to listen - only that there tended not to be one. What would have come as a surprise to a composer need not be taken as being equivalent to something he would have deplored.

So both of these criticisms, while they are useful and accurate, do not actually apply to a significant number of cases. In spite of them, it is still possible to find pieces whose composers had ideas about how they should sound - pieces of which modern performances sometimes seek to recreate the authentic sound. These first two criticisms are therefore not very damaging to either the purist or the missionary eclectic.

Points 3) and 4) are much more significant. That we cannot verify whether or not we have achieved the sound we were seeking cannot be countered satisfactorily. We may be able to verify that the instruments are contemporaneous, or even that our reproductions are accurate; but we still cannot be sure that the noises we get from them are the same. We do not know what the sound we are seeking, sounds like - if a band are trying to emulate, or improve upon the sound of other authenticists, then the sound they are seeking is not an historical one at all - it is a modern sound. Since we do not know what an early choir or ensemble sounded like, we cannot possibly claim that we can achieve, improve upon, or even approximate it. At best we can take the historical objects and blow them according to the manner described in historical documents - but even then it is does not follow that authentic sound is produced.

The only way to ensure that authentic sound is produced is to define authentic sound as being that which is produced by authentic instruments. It seems that some performers mean nothing more than this when they talk about authentic performance. For some performers, the question of verification is not at issue here, because the authentic sound is, by definition, the sound produced by period instruments, which is perceptibly different to the sounds produced by modern instruments. In this case authentic sound is little more than different sound. But our criticism is not directed at this account - we are concerned with those who believe that the use of old instruments gives them some special closeness to the situation of the composer. Godlovitch’s purists are such people, and again we find that our criticism is well aimed at them, but not so damaging to the missionary eclectics who might say that the attempt to recreate authentic sound is a way of interpreting differently, and that they are aware of the logical impossibility of knowing whether they had succeeded in doing so.

There might be some - purists or missionary eclectics - who might put up an intuitive defence. They might say that in spite of the logic, ‘common sense’ and intuition tell us that if we use the objects which were used, we are bound to get fairly close to the original sound, even if we cannot prove that we have done so. The sound - even if it were different - would not be significantly different, and would be sufficiently approximate to merit calling it ‘authentic’.

In reply to this, there would be three answers. Firstly, there is absolutely no foundation to this common-sense view. There remains a possibility that it is correct, but it cannot be defended on any grounds whatsoever. Performers can believe what they like, of course, but when what they say is utterly indefensible, there is a good reason for thinking again.

Secondly, improvement in the quality of authentic performance over the last twenty year or so indicates that a vast range of results ranging from the technically brilliant to the downright appalling is possible. Some of the early attempts on authentic instruments were unbearable, but as people have become more proficient, the results achieved are often of very high quality. Which of these is ‘bound to be fairly close’? There is no way of telling, and we cannot fall into the trap of supposing that the one we like best must be the most authentic. We cannot tell (which is the whole point), but it may be that today we can play the old instruments better than our forebears could. Thus the intuitive view is weak because it is not clear which authentic performance is taken as the paradigm.

The third answer is the same as point 4). Even if it could be said that authentic sound was reproducible - either intuitively or with logical foundation - the critic comes in with the observation that the significance of those sounds has changed irremediably:

"...as modern listeners can never hear with eighteenth-century ears, attempts at historical reconstruction (or so-called ‘authentic performance’) are preordained to failure. This argument maintains that, even if we could fully recreate eighteenth-century sounds, their effect on us would not be what it was on listeners two hundred years ago."

We are changed by the performances we hear, both of old and new works. The thrilling discords of Mozart’s ‘dissonance’ quartet do not sound so striking when we have heard Verklärte Nacht, and we cannot pretend that they do.

A further point can be made, either to underline this criticism; or it may be made by a missionary eclectic as a reply. This is that if performances and new works change us, we are actually saying that we are affected and changed by novelty and difference. If that is so, then it would be true that authentic performances also change us, and that we cannot hear a ‘vitalist’ performance of Mozart on modern instruments in the same way as we used to; because we have heard Hogwood or Gardiner play it on authentic instruments. The argument can be turned around, such that it is not a negative criticism of authentic performance, but a valid affirmation of it. We have seen how interpretation can be characterised as ‘doing something’ to a work, such that Taruskin claimed that those who aim to recreate the authentic sound are removing what has been ‘done’ by successive generations of performers. By the same token, authenticists have also ‘done’ something, when they get their period instruments out, and this ‘doing’ becomes part of the interpretation history of the music they play. The history of the performance of even Brahms’ and Vaughan-Williams’ works now include what the authenticists have done to them.

Thus, not only can the old significance of the sounds not be recreated, a new significance is being created each time the piece is performed or recorded. In virtue of this new significance, authentic performance takes its place in a collective tradition of interpretive stances.

2.4.4 Singularity of interpretation

At this stage, there is little more to add concerning this accusation, which we only make against the purists. In Part One we have already rejected the view that there can only be one interpretation of a piece, in virtue of Sibley’s theory of aesthetic qualities. We have also acknowledged that such a view is counter-intuitive (even its own advocates admit that), and is unhelpful to anyone who actually wants to perform anything. In the current debate, we have also seen that very few people actually hold this view, and that the value of doing so is not likely to be high. The purist is also the victim of almost all of the other charges we have explored. This being so, it has been more interesting and realistic to consider the position of Godlovitch’s missionary eclectic in response to these criticisms. But such a person does not advocate a singularity of correct performance, since a range of diversity was included in the definition of such a character. Thus, this charge does not concern the missionary eclectic at all since he or she has no argument with it.

Somebody might say that there could be a single interpretation which was of the greatest value, and that such an interpretation happens to be authentic. Thus a missionary eclectic might believe that an ideal performance could be achieved - but would not want to specify in advance how it would be achieved. Such performances tend only to be identified after or as they happen, and if they are recorded, they are taken as benchmarks for the future.

This is not the same as saying that a particular method, or style, or use of resources automatically leads to the best result. I am not denying that there can be a ‘best’ performance - only that such a performance could be reduced to a formula in terms of instrumentation or technique. Thus there is a point to be made - those who deny that authentic performance must lead to the best result, are not necessarily saying that there cannot be a perfect performance; nor that the composer’s intentions cannot be realised; nor that authentic instruments should not be used because they cannot produce the best result. The view is rather that authentic instruments could be employed in a performance which was considered to be ideal, and it could even be that the performance was ideal in virtue of their use. But an authentic performance which is the best performance is contingently so; never necessarily so, even if the authentic instruments are the key to the success of the performance. For on a separate occasion, their use could spell disaster.

2.4.5 The remaining points

Before summing up this section, we can consider those points which did not deal directly with what we have presented as the views of the coherent defenders of authenticity. These were criticisms 1 and 3. Point 1, put by Taruskin, characterised authenticity in terms of novelty; point 3 was Rosen’s concerning the rôle of recording. That Godlovitch did not address these criticisms means that they are either misdirected; or that the anti-authenticists have found a weakness which Godlovitch and his colleagues have not considered. A third possibility is that they are not so much criticisms as descriptions of the phenomena of authentic performance today. We shall now turn to these two points.

2.4.5.1 The novelty of authenticity

Taruskin’s central thesis is that authentic performance, far from being historical; is actually a manifestation of modern performance trends and techniques. Period performers should really claim to be doing no more than attempting a ‘new’ way of performing, whose main aim is to be different. The conclusion of this view, and the way in which Taruskin presents it, backed up by Leech-Wilkinson, is convincing as a commentary on today’s authentic performance.

So far, we have not come across anyone who denies Taruskin’s view. It is possible that the purist might deny it; but from what we have seen of the missionary eclectic, it is unlikely that he or she would want to deny it, since missionary eclectics tend to see themselves as offering one of many ways of interpreting, in which case the comment that their approach is novel might even be taken to be a compliment.

A purist might want to insist that the ‘authentic’ approach is historical, and is not modern at all. But we have already seen that historical verisimilitude is not attainable or verifiable - and Taruskin counters this possible objection himself. Much of Taruskin’s paper is devoted to a defeating of the purist case - which is what prompts Rosen to accuse him of ‘flogging dead horses’, because we have seen that the purist’s case is weak for other reasons too. Rosen believes that there are not very many purists to criticise, although he does accept that some of Taruskin’s dead horses have ‘signs of life’, by which he presumably means that there are vestiges of purism lurking around. But the majority of performers are not claiming to be historical, so once again we find that Taruskin has shot at the wrong people. Either he is not aware that many performers do not hold the view he criticises - and is confining his attack to purists (in which case we presume he has no argument with missionary eclectics); or, he is not able to make the distinction which separates the majority from those against whom his criticisms may be directed accurately. So the first point is, once again, that there are many authentic performers whose practices and views are not challenged by Taruskin’s critique.

Secondly, it is worth observing that authenticity is not unique in being novel. If the use of old instrument is novel, then it must be that the introduction of a new one also is. When the tuba was invented, or when Boehm devised a keyed flute, or when new forms of piano were introduced, performance practice changed and works sounded different. Philosophically speaking, the newness of authenticity is equivalent to the newness of any other performing style, particularly where the introduction of a new type of instrument is involved.

This is not of much consequence, but indicates that what Taruskin himself says about authenticity is not solely applicable to it, which itself shows that his argument does not say anything unique about authenticity; or set it up as a peculiar practice. What is distinctive about authenticity, would be that it is - or claims to be - historical. But we have seen that the claim to be historical is unfounded, and that very few people claim the historicity of anything more than the instruments they are using or copying. Taruskin’s critique does not distinguish authenticity from anything else. He singles it out for criticism, but proceeds to argue (presumably deliberately) that it is characteristic, not only of a modernist approach in its content; but also that by its nature as authentic performance, it functions in the same way as any other kind of performance. Authentic performance is equatable with modernist performance - but it is equatable with any other kind of performance too. The problem though, is that he has distinguished it as a particular phenomenon, in order to argue that it is no different, which - if he is right, suggests that he should, or could not have separated it in the first place.

This may account for his, and our, difficulty in defining authentic performance (it will be noticed that I have not attempted to offer a definition). Taruskin listed a few alternative terms, dismissing them, and also rejecting ‘authenticity’ itself, because of its negative antonym. For convenience we may take the view that authentic performance involves old instruments, but this does not satisfy those who want to say that period performance style and technique are involved. The point here is that we may not be able to define authenticity, because as Nikolaus Harnoncourt believes - it simply does not exist:

But first of all, you know, I began my career as a perfectly ‘normal’ musician. I have a completely different attitude to the whole historical area than that which others later attributed to me. When I hear the word authenticity it drives me up the wall. Authenticity simply doesn’t exist. I’ve read in several record notes that they are reconstructing exactly the original forces, eight first violins, seven seconds and so on. It is so idiotic. When I read something like that I feel it has absolutely nothing to do with music."

If Harnoncourt is right, then ‘authentic performance’ is simply performance. Taruskin’s attempt to isolate it from the performing tradition of which he himself says it is a part, means that his account separates it in order to show that it is not separable. Either he is wrong - it can be separated, and that it can be is proved as much by his having done so); or he has contradicted himself. Many people have separated authentic performance from performance though, although our experience has been that it is a messy task, authenticity cannot be isolated completely because of the ‘pastness of the present and the presence of the past’.

To give Taruskin credit though, he seems to be saying that authenticity can be separated out from the modern context; but that it should not be. But his argument about why it should not be seems to hinge on why it cannot be. This is strange, because if Taruskin believed that the separation really cannot be done, then the question as to whether it should be is an empty one. It is likely that Taruskin believes that there are people who think it can be done, and that it is his task to point out that they are wrong. But such people - whom we have called purists - are few and far-between; and their being in the minority consigns Taruskin to the sidelines of the debate, no matter how clever and interesting is his defunctive equestrianism.

Thus, it may be said by a missionary eclectic that authenticity makes no claim other than to be part of the range of performance possibilities today. Being novel is not a problem for authenticity, or for any other kind of performance practice, and it is certainly not distinctive of authentic performance. Novelty is part of the aim of any performer - not necessarily for its own sake, but in the pursuit of freshness and vitality. Authenticity seeks the same goals as any other kind of performance, which suggests that it should not be isolated without considerable care. When the outcome of careful isolation is that authenticity represents the same enterprise as that from which it was isolated, the conclusion must be that its isolation is invalid, because in the process of separating something from the whole, in order to prove that it is part of the whole which cannot be separated, one has inadvertently proved either the separation, or the thesis to be incorrect, but without being able to tell which.

2.4.5.2 Recording and technology

Our final criticism or observation was Rosen’s - that the authenticity movement depends heavily upon the technology and financial support of the recording industry. This itself is not a criticism, or fault - musical enterprises and practices have been supported and sponsored by patrons throughout Western history. But Rosen does criticise the tendency to render performances inauthentic by recording them at a high volume, thereby offering the home listener a product which does not produce an experience of equivalent dimensions to the performance recorded in the past, let alone to the putative performances of centuries ago.

The purists can hardly say in response that they have the benefit of recording to present his privileged approach, because as Rosen rightly points out, the presentation of that approach makes a nonsense of what they were trying to achieve. Having reduced the work to its original dimensions, so to speak, the recording engineer and home listener blow it out of proportion again. If one can be authentic (or if one intends to try), then the attempt should be made at every stage of production. Authentic sound cannot be authentic if a record producer has doubled the volume, or realigned the balance of instruments. The criticism is that recordings aofnd authentic performance - which may seem appropriate, are actually inconsistent with the aims of authentic performance.

Purists and missionary eclectics alike might respond that they cannot prevent listeners from turning the volume up, beyond a level that the instrument(s) would have achieved in one’s own home. This seems a fair defence, because we can hardly blame someone for what other people do with what they have produced. After Hiroshima, we did not blame Oppenheimer for having developed the technology which was later used to make the nuclear bomb (even though he seemed to blame himself). On a different scale, we do not accuse authentic performers of making inauthentic recordings, because someone plays them too loud. It seems more reasonable to see this aspect of Rosen’s point as more of an ironic observation than a serious challenge to period performance practice.

This being so, we would also have to observe that this would be true for all music written before the recording era, and more much written after. The exception is perhaps popular music, which is written with the intention that it be recorded, and that certain techniques of recording and certain habits of reproduction be employed. Rosen’s point - like Taruskin’s about novelty - is not distinctive of authentic performance, although it is true that the irony is stronger when the aim is to produce historical sounds.

Some performers might say that their performances are helped by the recording, and that they sound better when recorded and played at unrealistic levels. This may be a matter of opinion, but we cannot ignore the possibility that a ‘louder than life’ approach gives the recorded performance and extra clarity, or makes it more exciting acoustically. The purists would object to that , but it is not them we are concerned with in this case. If one is aiming for the best result, there is always the possibility that enhanced ‘authentic’ sound produces it for some cases. Many modern performances of Bach’s keyboard concerti employ a modern piano, because a harpsichord does not make enough noise with a larger orchestra. There is a case for using a larger orchestra with harpsichord, and then increasing the level of the harpsichord, such that the result is pleasing and well-balanced, even if it could have been achieved ‘live’.

Finally, there is a feature of recorded performances which has not been touched upon by Rosen. An authentic performance, like any performance, is usually an event. It begins at one point in time and finishes at another moment, later. But most recording practice dictates that the ‘performance’ committed to tape, is not a recording of an event. It is not my concern here to go into great detail and discuss whether a performance must be of an event to be such; but it is to be observed that a recording and a performance differ in this fundamental way. Neither does this observation pertain particularly to authentic performance and authentic recordings. But there is another irony here, since the attempt to record authentic performances - while it may render them inauthentic, it may also be said that it also renders them non-performances. A patchwork of takes is not a performance, if a performance must be an event. So authentic performers - of all performers (to whom this point is of interest) - even if they successfully navigate the philosophical minefield of authenticity, may inadvertently blow themselves to pieces as soon as they commit themselves to disc. The only way to avoid this, is either to show that performances need not be events; or to ensure that all recordings released are of ‘live’ performances, produced in a single ‘take’.

The issues raised by recording are mostly ironical, rather than harsh criticisms. Nevertheless, they are worth mentioning, because authentic performance has so often been associated with the recording industry. The view that authentic performers should have nothing to do with recordings is a little extreme, especially when the relevant technology can enhance the result. If the current procedures of multi-tracking and re-taking are employed, then there may be a case for all musicians to be wary - although this is by no means established, and the value of the compact disc, tape and record to our musical lives is plain to see, hence the suggestion that recording is inappropriate would contradict almost everyone’s experience.

2.5 The verdict

Having looked closely at the criticisms which Rosen and Taruskin have levelled; and at the way in which purists and missionary eclectics might respond, we see that the purists are under heavy fire, but that the missionary eclectics remain largely unaffected by the points made against authentic performers.

2.5.1 The purist

The purists have been criticised on almost every point. We have seen from Part One that they cannot argue that their authentic approach has any kind of privileged status; nor that their approach is the singular best because it respects the composer’s intentions best; or reproduces the sounds or stylistic practices of the past.

Frank Sibley’s ‘aesthetic concepts’ show us that we cannot prescribe or predict that an authentic performance will either be the best - or even that a certain approach will necessarily be authentic, if we take ‘authentic’ to be a value term, like ‘beautiful’, or an aesthetic term such as ‘graceful’. In neither case can the result of a particular approach or action be stipulated in advance, in which case the purist’s claim is flawed from the start. The purist can hardly say that his or her approach happens to produce the best result, because to do so must acknowledge that another approach might produce a best result (even if it does not actually do so, there must be an acknowledgement that it could); and that would be the missionary eclectic’s position.

On a similar note, the purist must be wary of trying to ‘strip away’ the crust of later performance practice from the tradition through which the piece has been handed on. The idea that the original work needs to be rediscovered, or ‘cleaned up’ suggests that the human interaction with the work that has gone on since its creation, is of no consequence, and is to be rejected. Purists might deny that their approach amounts to a rejection of the modern world; but a ‘back to Bach’ mentality could well be dehumanising and negatively critical of other performers today.

Nor can purists say that their approach best respects the composer’s intentions. They are not affected by the charge that they must respect the composer’s poor judgement as well as his good judgement. The purist is well-advised to be wary of confusing what the composer wanted with what he got, especially where the attempt to reproduce authentic sounds is made. But it is possible not to fall into this trap. Similarly, it is not a major problem for the purist if someone points out that some pieces were not meant to be ‘performed’ in public at all. For just as one need not respect the apparent intention to have a performer adhere to practices long defunct, one need not refrain from performing pieces which weren’t directly intended for large audiences.

Purists cannot claim that they can reproduce a performance which lacks interpretation - which does no more than play the music, for two reasons. Firstly, it is not possible to play without exercising some degree of choice in matters not specified by the composer’s manuscript. Secondly, the distinction between playing the notes and interpreting them is a false one, supported by the popular conception that one can learn technique and execution before ‘putting the music in’. Admittedly, a rendition which attempts to concentrate on technical execution rather than interpretive gesture will probably have a limited artistic value; but one cannot remove the hand of the performer, nor can it become solely involved with the transmission of the artist’s intentions.

In preparing performative interpretations the purist might try to be accurate in recreating the original conditions or context of a first or early, or ideal early performance. Such an achievement could not be verified, even if we could decide which early performance would make a good paradigm for emulation. But since such conditions have never been fully specified - nor could they be - a completely contextual performance is not possible, and so purists who strive for it must come to terms with the futility of their quest.

If a purist seeks to reproduce the sounds of the past, he or she must realise that the attempt cannot be verified, and the intuitive idea that something approximate must result from using old instruments is not philosophically respectable. For the significance of those sounds - even if they could be reproduced - will have changed because we have heard more modern music since; and live in a world with different values and ideas. If we cannot be authentic hearers, then we cannot tell if we are being authentic players.

The idea that a purist might have - that an authentic performance can be fixed onto tape is attractive; but ironically, the process of recording changes rather than fixes. The business of recording raises questions about the nature of the performed result, and a tape recorder is no authentic instrument.

At the end of Part One we had reason to believe that the purist’s case was not a particularly strong one, and a study of the criticisms of authentic performance in general has revealed that much of that criticism is directed at Godlovitch’s purists. To a certain extent the purists are represented as taking the extreme position - and for some criticisms we have good reason to doubt that anyone takes the extreme view (for instance, it is unlikely that anyone actually believes that the conditions of a first performance can be perfectly duplicated). But our critics have warned off anyone tempted to such views and it seems fair to say that in doing so they have kept some performers away from the extremes and made them think about the implications and possibilities of what they are doing. It may even be that the harshness of the criticism of the purists has ensured that most modern authentic performers are best described as ‘missionary eclectics’.

2.5.2 The missionary eclectic

If our debate rules out the position of the purist, then we are left with the missionary eclectic. We are also left with the hodiernists, but we need only return to them if we find that the position of the missionary eclectic is untenable. From what we have seen, this is not so. The missionary eclectic has not been damaged very much by the criticisms we have discussed, and the case for this position seems to have been bolstered, not least by the implications of Taruskin’s views.

Missionary eclectics affirm diversity of interpretation within a performing tradition, which leaves them unaffected by four of the criticisms we have discussed. They also accept that complete contextuality is not possible, and so, where historical instruments of old performing techniques are employed, they are used in full awareness of their limitations; and no sweeping claims are made. Rather the resulting performance is offered up to take its place in the range of performances experienced by listeners, and to be judged for its freshness, vitality, novelty and sensitivity.

Missionary eclectics do not eschew the intentions of the artist - they tend to respect them. There is an awareness that intentional information is never complete; but also a tendency to look at the music, rather than outside it, for appropriate evidence of how the composer intended it to be performed, and what kind of effect he had in mind. Missionary eclectics, like purists, have a great respect for the score, and may well believe that the verbal instructions, as well as the notated pitches, serve to express the intentions of the composer. Performers can attempt to adhere to these, and for as long as their attempt to adhere to them is recognised to be such by the listeners, then their endeavour will serve as a performance of the particular work, to be judged on its own merits and compared with ‘benchmark’ recordings and performances. In this way, authentic performances take their position in the wider spectrum of performance approaches.

Some of the stranger intentions or expectations of the composer do not present a problem for missionary eclectics. Where a composer’s ideas about his own work are flawed, or out of date, then scope for improvement or editorial revision is acknowledged. Similarly, there is no need to keep up with outdated practices for beating time or keeping in tune, when it is clear that such practices are no longer necessary. Nor does a missionary eclectic have a problem where the performance of pieces not originally planned for public performance is concerned. For if the work is valuable and can be performed, then this is a good reason to perform, because the aim to be authentic is secondary to the aim to produce something of value.

Finally, the missionary eclectic is not challenged by the charge that authentic performance is modernistic performance. Concerning the charge itself, he or she would probably deny that their approach was mechanistic and non-interpretive, since there is usually an attempt to produce something of more than historical value. But by implication, Taruskin’s point admits a separation of authentic performance from other kinds of performance. Missionary eclectics would want to deny this - for them authentic performance is nothing peculiar, it is performance, and makes no special, unique claim for authority or value. In this much, the place of authentic performance in the general range and history of performance is affirmed. For inasmuch as every new way of performing of Mozart’s Requiem has changed us since its composition, an authentic performance can also make a difference. The ability of authentic performance to make such a difference is to be affirmed, not only because it demonstrates that authentic performance is performance but also because the effect of authentic performances can be - and have been - very positive and valuable. Authentic performances have changed us, and in that they have done what other kinds of performance have been doing for hundreds of years. The purist is not able to admit this very easily; but the missionary eclectic is able to relish it.

In our survey then, the missionary eclectic not only defeats the purist, and withstands the criticism of Taruskin and Rosen, but emerges in a stronger position than Godlovitch originally proposed.

3.1 Authentic performance and authenticity

It has not been my purpose to define or account for the phenomenon of authentic performance. Rather I have sought to discover whether the modern performance practice, known to its advocates as ‘authentic performance’ is what it claims to be; and whether or not it is a viable approach to musical performance. I began by offering a broad description of what is meant by ‘authentic performance’, and pointed out that although there is much literature on the subject - especially in journals - very little approaches the subject from a philosophical point of view. Hardly any offers a balanced account of the pros and cons, or arguments in favour or against any of the arguments or assumptions which we have met in the course of this enquiry. There is, to my knowledge, no book on the subject.

It has therefore been my project to discuss how one might defend the playing of old instruments according to period practice. Then I have sought arguments which challenge the beliefs or assumptions upon which such a defence may be built. Having worked out the former assumptions and presented the latter criticisms of authentic performance I endeavoured to join them in a debate among three characters - the critic, the purist and the missionary eclectic. Both the missionary eclectic and the purist advocate authentic performance, but the purist typically suggests that authentic performance should be employed, whereas the missionary eclectic is more likely to say that it may be employed. The purist is an advocate of the view that there is a single interpretation which does the piece and the composer’s intentions in respect of it, the greatest justice; whereas the missionary eclectic believes that there is more than one way to perform a work, and welcomes the contribution of modern approaches too. The aim of the missionary eclectic is to produce great performances, rather than authentic ones.

In the course of the debate it was noticed that much of the criticism that has, or could be made, is directed at the purist. The critics may not be aware of this, but much of their argument reflects badly on purists, but does not damage the philosophy of the missionary eclectic. Sometimes the purist’s critic inadvertently supports the missionary eclectic simultaneously. The missionary eclectic’s position is tenable, the purist’s untenable. Thus at the end of our enquiry we discover that missionary eclectics have philosophical blessing on their approach.

While it is not strictly within the scope of our enquiry, it is appropriate to discuss that approach in a little more detail. If the missionary eclectic seeks to achieve something by playing on old instruments, we can wonder what it is, and how it is achieved. If authenticity, as Taruskin put it, is:

"...knowing what you mean and whence comes that knowledge. And more than that, even, authenticity is knowing what you are, and acting in accordance with that knowledge. It is what Rousseau called a ‘sentiment of being’ that is independent of the values, opinions and demands of others."

...then it is worth spending a few words on the connection between authentic performance and authenticity. For I have spoken of ‘authentic performance’ throughout, but not of ‘authenticity’, or of ‘being authentic’. Taruskin drew a distinction between ‘authentic’ and ‘authentistic’ performance, which I have not adopted. I have been discussing authentic performance, not authenticity. In conclusion I hope to show the usefulness of this distinction, and suggest what ‘authenticity’ might be.

‘Authentic performance’ is not easy to define. I have already suggested that this may be because it is not actually separable from ‘performance’. Taruskin segregated it, but concluded that it was not a unique phenomenon, and implied, along with Harnoncourt and others that as such it probably does not exist as distinct from modernist performance, or performance in general. Even now I do not wish to ‘define’ authentic performance, because it is a slippery concept understood differently by those who refer to it. On the other hand, most of us have an idea what is being spoken of when we read the phrase on a record sleeve or concert poster. We have expectations and associations of old instruments, smaller ensembles, lower pitch, less vibrato, more punctuated rhythm, faster tempi, and thinner sound. We may see these as advantageous, or may have different, negative expectations involving poorer tuning and a plain, unflamboyant approach which may not be to our taste. This is not a definition of authentic performance, but I hope to have typified it.

To some people the word ‘authenticity’ will have the same meaning. Where the two terms are used interchangeably, I fear that a great deal of confusion may arise. As I said, I have been concerned with authentic performance, not authenticity. Authenticity, as Taruskin implies, is something quite different, and whether or not the achievers of authenticity employ period instruments, strict rhythms or quicker tempi makes no difference to the authenticity of their performance.

We may say, with Taruskin, that being authentic is about "knowing what you are doing and why". It is about making decisions and choices - not only concerning the composer’s intentions; the authority of the score; or the performance practices of the day - but about whether those factors are to be taken into account in the first place. Authentic performers who decline to make choices when performing may well produce authentic performances, but it is very unlikely that they will produce a performance which possesses the quality of authenticity.

"But as long as we know what we do want and what we do not want, and act upon that knowledge, we have values and not dirt. We have authenticity."

Some authentic performers would disagree - not so much that authenticity and authentic performance are to be distinguished, but that certain forms of performance preclude authenticity. Christopher Hogwood is a good example:

"The performance of Messiah here recorded attempts to recreate the context, sound and style of a particular performance given under Handel’s own supervision; the special considerations involved are set out below. At the same time it is of course intended to be a joyous and vital act of music making, to which end all historical and musicological investigation is directed. Only those who are inadequate scholars or inadequate performers (or both) will find any conflict between these two complementary aims."

Hogwood no doubt believes that the purist who attempts to recreate the context and sound, can also be authentic. He can know what he wants to do and what he does not want to do, of course, and this seems to suggest that he can be authentic. But our enquiry has shown that what one wants to do must also be related to what one can do, and if he is wanting to do something that is impossible, then Taruskin’s idea that one must know what one is doing is not being recognised. Because ultimately Hogwood is failing to achieve what he sets out to do. Intending to make a joyful noise is not tantamount to making informed choices, admirable as that aim is. But I wonder who performs Messiah without intending to produce a vital and joyous result. Hogwood’s final comment is rather rude, suggesting that anyone who disagrees with him is incompetent. He is also criticising one of his most respected authentic performing colleagues: John Eliot Gardiner uses ‘period instruments’ of course, but makes no wild claims or moralising justifications for doing so:

"...how in our changed circumstances, even with the benefit of scholarly research into Bach’s own performance practice, do we begin to do justice to this epic work which was written, of course, for divine service, not concert performance (let alone a recording!). To seek to reproduce the exact performing conditions that Bach had at his disposal in 1727 or 1736 is a chimera: it is both figment and fantasy...

Just as important as the performer’s need to investigate all that there is to be inferred from Bach’s own practice is his and the listener’s need to come to terms with the rhetorical language of Bach’s Passion music and to understand its theological purpose.

Gardiner seems to be more aware of the distinction between authentic performance and authenticity, and it seems fair to presume that he strives for authenticity when performing with the Orchestre de l’opera de Lyon as well as with the Monteverdi Choir, The English Baroque Soloists or the Orchestre Révolutionnaire et Romantique. No doubt Gardiner aims for what Andrew Porter described as "a sense of great events moving toward far off conclusions", a sense that the performance is contributing to the life experience of those participating as performers or listeners. Such performances are those which change us - not so much that ‘do something for us’, but those which do something to us. Great, or truly authentic performances are not predictable, or even repeatable - whatever kind of quality ‘greatness’ is, we cannot produce a formula or method for achieving it.

3.2.1 Greatness

Greatness is not the same as authenticity, since from what we have seen, it seems possible to produce an authentic performance which is not necessarily judged to be great. A performer may know very well what he or she is doing, and why, and have a very good idea of what the desired result is. But the result need not be great. John Eliot Gardiner and Nikolaus Harnoncourt may know what they are doing and why, and make no extreme claims for their use of period instruments; but their performances are rarely considered great. If any of their efforts do produce an outstanding performance, then it is not because they are authentic performances, nor because they are produced authentically. Rather it would be because their interpretation changes us, moves us, or carries us on to a new and valuable experience of the work and its world. Even if one was aware of all the philosophy, musicology and history a great performance need not ensue; although such awareness - and the acting upon it - does tend to produce authenticity.

This is to a suggest a much more important distinction than that between authentic performance and authenticity - one between authenticity and greatness. For authenticity does not demand authentic performance; but great performance demands authenticity as a minimum. For one could hardly produce a great performance by accident, great performances tend only to be created by interpreters of a certain stature. Authenticity is not detectable whereas greatness is something recognisable, even if it cannot be prescribed or accounted for in words. It is very difficult to generalise about greatness - if only because great performances are - perhaps necessarily - rare. A great performance is one that stands out from the rest, it is distinct and distinguished in virtue of particular qualities which it possesses.

We cannot predict or define what makes a performance great. Andrew Porter got close when he wrote of a "sense of great events moving toward far off conclusions". A great performance has direction, a motive, or underlying logic. Sadly, the great performances and recordings which we hear are diminishing proportionally, in spite (or because) of the thousands of recordings and concerts made annually. The authentic performance movement has made us review the ways in which we actually play the music; but in doing so has challenged the more significant practice of interpretation. Now that authentic performance has been assimilated and has found its rôle, and we have learned useful things, we should re assimilate our desire to convey a sense of greatness through performance. It may be the case that authentic performance and interpretative greatness do not go hand in hand, and it is a pity that no-one has yet been willing or able to try. But then, there are not very many conductors today who qualify to make the attempt, although it would be interesting to hear Guilini conducting the London Classical Players. If he were to do so, we would probably find that any interpretative genius he brought to it would transcend the incidental qualities gained through the use of historic instruments.

But if we are in search of ‘greatness’, is there anything that can be said about that which we seek? A theory such as Goodman’s, which emphasises ‘correct’ performance, is interested more in whether performance P is of work W, than whether or not it is a good or great one. Greatness is not discussed by many aestheticians - perhaps because it is such a difficult area. Peter Kivy has written about what he calls ‘profundity’ in music, but even he is careful to distinguish between profundity and greatness:

"Now in saying that I have, so far, failed to find any rational justification for calling musical works profound, I do not want to be misunderstood as denying that, at the present time, I have rational grounds for thinking certain musical works great works of art. I am not using the word ‘profound’ as synonymous with the word ‘great’ or any other word like it. I think all but not only profound musical works are great musical works. And I think we do have rational grounds for thinking some musical works great, and some greater than others."

Kivy is talking about works, rather than performances, but if we are able to say something about great works, we may have a key to unlock great performances. Kivy though, does not provide it - he prefers to talk about ‘profundity’, and does not discuss what those ‘rational grounds’ for calling some works greater than others might be.

3.2.2 Carl Dahlhaus

Kivy thus begs the question about what makes a work great. One of the few writers to consider the concept of a great work was Carl Dahlhaus. In concluding his book on the esthetics of music (sic) he discuss what may be called the ‘chaos of judgement’. The apparent confusion caused by widespread dispute over value can be remedied, either because the examples under discussion are inappropriate, in which case there is no real dispute; or because there is no tradition on which critics may draw, when entering debate.

Where a disagreement is apparent, there is often a "mean and impertinent" opinion. In fact, divergent opinions proceed from different premises or assumptions, and so what seems to be a dialogue leading to disagreement, is not a dialogue at all. If a common basis for discussion could be arrived at, then these opposing views might complement rather than contradict one another.

Thirdly, Dahlhaus counters the prevailing view that the history of music criticism constitutes nothing more than the history of prejudice and polemic. The continued debate is healthy and fruitful - we may have rejected ‘vitalist’ criticism - as Taruskin might call it - but we would be impoverished if we ignored the insights which Romantic criticism has given us:

"Our incapacity to make some of the discoveries of the earlier centuries does not preclude our holding onto discoveries once achieved. We need not share the presuppositions of past epochs, nor worship their idols, in order to participate in the insights that grew out of them."

He also makes the point, familiar to many philosophers, that even though the inability of critics to agree about judgements is much cited, is not actually so common. Even where there is disagreement, a great deal of communality must precede it, concerning whether the object in question is art at all, and also an awareness of the many issues of aesthetic principle which may be drawn upon in conducting critical debate. Different points of view respect, and are aware of, one another’s perspective of judgement. Thus:

"Some work of art flawed from the point of view of perfection may be significant from the point of view of greatness. And nothing has exposed esthetics to general contempt more than the strained effort, dictated by insistence on system, to gather all possible specifications of works of art around the central idea of beauty, or even to deduce them all from this idea."

When a culture or age is determined to live by systems, this kind of attempted synthesis is to be expected - and with our picture of Stravinskyesque authentic performance, we may have an example of its consequences - but Dahlhaus says that there can be no hierarchy of aesthetic ideas, rather they coexist and cannot be subordinated under an overarching idea, such as ‘beauty’, (or ‘authenticity, for that matter).

This leads Dahlhaus into the territory of greatness, which he distinguishes from ‘perfection’. He proposes that a ‘perfect’ work stands in its own special world, and that any appeal to "biographical or historical reminiscence" detracts from it. Musical greatness, on the other hand, strives beyond its own limits, and is more than just ‘pure sounding’. Great music, which is open to extra-aesthetic considerations, also lends itself to the idea that there is a ‘speaker’ behind what it heard:

"With works of Beethoven, Wagner, or Mahler, it is quite hard, and would probably be unrewarding, to let esthetic pedantry prevent thinking of the composer’s personality... the idea of a music history without names is rooted in classicism, whose esthetics circles around the idea of perfection, counterpole to the idea of greatness."

Greatness in music is not independent of extra-musical factors. Pieces that have a narrow scope, such as two-part inventions, or strophic songs, can hardly be great. Larger forms, especially those which demand grand-scale or large resources, demand a rare and ‘almost despotic mastery’. While perfect music is not hard to come by, great music is associated with the concepts of monumentality and difficulty. Great music is not immediately accessible, and runs the risk of becoming sterile and monstrous.

3.3 Great performance

Great works merit great performances. Even if we have been helped by Dahlhaus’ brief comments about great works, we still need to relate them to performances, which are our concern here. Of any decent performance we would expect the right notes; a sense of musical direction, form and structure; a concentration on the forming of each note (a concept which may make more sense to wind and string players, but which can nevertheless be applied to keyboard playing); and an awareness that in the performance something is happening. The performer presides over an audience - he or she is the agent of their experience - and as Urmson has shown us - has duties both to them and to the composer. Just as a great work transcends the sound, drawing in the composer’s personality, a performance can draw in the personality of the performer. Indeed, it is what the performer ‘gives’ to the work in performance, that determines whether or not it should be called ‘great’.

As a performer, I must pay attention to the piece being performed. Whether the work itself is of value is to be distinguished. Hence I may perform great work GW, or mediocre work MW, and yet in doing so, I may take the same performer’s attitude to both. I will get to know the piece, and think of it in terms of some kind of journey - I will try and convey to the audience that during the performance I am attempting to take them from one place to another; and I will have concentrated intensely on the individual details, as I lead along the path the music traces. But it is not me who has made the path, which itself may be rough or boring. The work may not be a great one, but a performer may still try to make even a well-worn, or uninspiring path interesting. And it is probably true that the staple of our musical experience does not consist of great performances, nor of great works. Most of the time we experience good performances of good works. Sometimes those performances are so good that we want to call them ‘perfect’. But we do not need, or even want, to experience greatness all the time. Rossini’s overtures may not be great works, but we are more than content to attend performances of them, because we do not demand that our life be changed every time we go to a concert. But we do expect the performer(s)’ execution and interpretation to be of a certain standard.

Here we are reminded of Dahlhaus’ distinction between perfection and greatness - for it becomes clear that there can be a ‘perfect’ performance, where these standards of execution are met, but a ‘great’ performance is quite something else, and may even be hindered by perfect execution. I was once told a story about Artur Rubenstein, who - when asked to comment on a young player’s performance - commented that there were "too many right notes". Greatness in performance, as in aesthetic theory, is to be distinguished from technical perfection.

The perfection may even extend to the ability of a performer to interpret. Most of today’s performers (authentic or not), know what they want to do, and why, and offer it, not only as the work of composer X, but also as the performative interpretation of performer(s) P. Even at the ‘everyday’ level, the performer makes some claim of ownership, both of the particular performance, and of the PI it instantiates. But this is not sufficient to greatness, even if the execution and interpretation are perfect, or ‘authentic’, or both.

But what we have observed about performances in general, must also be true for great ones. But while there can be an everyday performance EP of either mediocre work MW or great work GW, this does not seem to hold for great performances. For it is counter-intuitive to suggest that there can be a great performance GP of work MW. Even if we think about perfect works, which are not great (and I think we must) - PW, I still want to resist the idea that there can be GPs of PWs. Great performances are confined to great works.

There can be perfect performances of great works, or even bad performances, but neither does justice to the ‘greatness’ of the work. A great performance must somehow feed off and present the greatness of the work, but more than that, it must also possess the quality of greatness in its own right. By this, I mean that the greatness is in one sense dependent upon the greatness of the work, without which it could not exist, but also that the great performance also has qualities which other performances of the great work lack. For example, we may take J.S. Bach’s St. Matthew Passion to be a great work; and may conclude that John Eliot Gardiner’s recording is ‘perfect’. We may also say that it is ‘authentic’, in that he carried out extensive research, and concluded that the work divides into sections, not unlike the ‘Stations of the Cross’. The result is novel, refreshing, and it is clear - whether one reads the liner notes or not - that Gardiner has put something of himself into the recording, and has done so, knowing exactly what he was doing and intended to produce. We can reasonably assume that what we hear on disc is pretty much what Gardiner intended us to hear. Gardiner’s version is also an authentic performance, in that it employs period resources. None of this adds up to greatness however, nor could it. Gardiner has contemplated the extra-musical, in dealing with the spiritual and meditative dimensions, as is appropriate to the greatness of the work. But the greatness - or lack of it - in the performance is determined only by Gardiner, not by the work he chooses. Yet it is also the case that in performing the suite from Rameau’s Dardanus, he prevented himself from producing a great performance, because the work itself is not great. Greatness of performance is something which must be compounded with greatness of work.

Otto Klemperer’s recording of the St. Matthew Passion, with Peter Pears, Dietrich Fischer-Dieskau, Elizabeth Schwarzkopf and the Philharmonia Choir and Orchestra is not an authentic performance, nor is it to be considered ‘perfect’ where points of execution are concerned. Dating back to 1962, it still conveys a intensity of devotion that directs us away from the technicalities of the production, to something far beyond. Questions about ornamentation and tempo raised recently by musicologists seem irrelevant in the presence of a musical act that takes hold of Bach’s theology and reverence and transmutes it with the alliance of his own intense feeling. Klemperer’s interpretation is more than a performance - it is a response, which transforms its object.

A great performance points to something beyond the basic experience of accurate rendition, possibly revealing something of the performer’s history and biography, and also engages with the greatness of the work performed. This greatness of the work is characterised in a similar way - the work strives for something extra-musical, and it is appropriate to think of this transcendence in terms of biographical and historical features. Thus there is an encounter or engagement of greatness where these two ‘transcendences’ meet. For both great works and great performances it is appropriate to enquire of biographical and historical issues, and, due to the nature of greatness, we may well find that such an enquiry is answered by reference to the composer-in-the-work; or the performer-in-the-performance, both of whom need not be too difficult to discover, because of the intensity of personality which great works and performances exhibit.

If only great works can have great performances, then it is worth noting that some authentic performances cannot be great, even if they can be authentic. There is little of Rameau’s music, for instance, which is usually considered ‘great’ - yet it is music such as his that is often treated to authentic performance. Such an authentic performance may also be authentic, if the performers approach it with a certain ‘sentiment of being’ as Rousseau would call it. But because the works chosen by many authentic performers are not of sufficient stature, they cannot hope for a great performance.

Furthermore, the performers themselves may not be of sufficient stature, and may only be capable of producing an everyday performance, at best a perfect one. They may, for instance, have no personality to speak of - or rather - no passion or intensity of feeling which they can bring to their work. A performer who, for instance, cannot relate to Bach’s theological and reverential world, will be at a disadvantage. To be great, it is likely that the performers must be able to augment that world, with some passion of their own. If this is right, then it may be that greatness in authentic performance is hindered, not only by the music performed, but by the people performing it.

3.4 The power of greatness

Having talked about great works, and great performers, it is appropriate to mention the effect on the third party - the listener. For I suspect that there are some performances of certain works which change us. They have a sort of sacramental quality - able to effect or reflect change, but in a way that cannot be unpacked or explained completely. They are able, as Philip Sheldrake would put it, to "point beyond (themselves) to a profound unity between persons". It may even be that such performances do not appear to be aesthetically distinct from many others, to many people. But even without being able to pin down what it is that enables some performances to change us, it is helpful to isolate that idea, and contrast it with the other experiences of music which we have on an everyday basis.

I have one further idea which may shed a small shaft of light on greatness, with which I wish to leave the reader before closing. In the course of our lives, we meet many people, and share all kinds of experiences with them. The people we encounter, and especially the people we love, change us, and are changed by us. This is because they matter to us, and are in some way part of us - there is a sense in which friends ‘own’ each other in virtue of the experiences they have shared. Sometimes an encounter or experience can have a lasting effect, it can change a person - not necessarily permanently - and alter their perception of themselves, or of the world. This is particularly true if the encounter has a strong or distinctive emotional dimension.

I have suggested earlier that the relationship we have with artworks is similar to the relationships we have with other people, and I want to pursue this a little further. For it may be that if we value works of art rather as we value people; the effect they have on us may also be comparable. A striking painting, or profound novel may change our lives, and this ability to change lives may be widespread - that is, a particular novel may influence the lives of many people because of its ability to speak truly about and to the human condition. The works of Dickens, Austen, Lawrence and Golding are examples here. Other works only influence a handful of people, or a certain type of person, because they deal with issues that only concern particular people at a certain time. Some novels only speak to a particular generation.

Without going into the philosophy of music and the emotions, we can acknowledge extensive discussion in this area, which itself testifies to the relevance and significance of the relationship between music and emotions. If it is true that a musical performance can move us to an experience, which can at least be likened to that of an emotional experience, then we may suppose that if that (quasi)-emotional experience is powerful, it can change a person, either because it is an emotional experience, in which case it can change us like any other emotion can; or that if it is not a genuine emotional experience - in that it lacks an object - or for whatever other reason - then it still might share the ability to make people feel different, or changed in some way. Many people’s experiences of concerts would support this kind of view.

We should be wary, of course, and remember that many people’s emotional experiences brought on in response to music, are not directly related to the music, or the performance, and so would not be directly relevant to this discussion. For many people, a piece evokes personally associative emotions which have been imposed upon the experience by themselves, albeit unintentionally. We must also acknowledge that at certain times we are more susceptible to emotional arousal than at others - such as after a bereavement. But if we can be aware of these factors, and eliminate them, we still find performances which affect us deeply and which we call great. I have suggested that on such occasions, the great performance is some kind of response to the greatness of the work. The listener who benefits from a great performance, recognises this relationship, and is affected by that recognition.

If we remember the personalistic analogy, we may think of those times when our friends reveal or share things which are deeply personal, or which matter a great deal to them or us. Such occasions, on which our engagement with them is intense, or passionate, or moving, are rare, and unrepeatable, and even if we can describe the content and effect of the experience, we cannot predict or engineer it.

We have already seen how we need have no philosophical fears about engaging with a work of art on an extra-artistic level. Dahlhaus would be inclined to agree - at least where great works are concerned - they almost demand us to look beyond them to the circumstances of their creation. If these works appeal to us, it is likely that we should want to do this anyway. Here the two theories coincide - we can go beyond these works; and their stature serves to encourage us to do so.

If this is so, then we must conclude that a desire to interpret only in terms of the aesthetic object, hinders great performance. At best, perfect performances could be achieved. But without great performances, we would lose the concept of a great work; and our great performers would be seen as no better than sentimental buskers; and we would (officially) be deprived of those rare events when we are changed by the recognition of a deeply personal dialogue talking place between the performer and the composer-in-the-work. I doubt that anyone can deprive us of such experiences, not leastly because it is not possible to confine our interpretation to aesthetic and extra-aesthetic. It is far more helpful to think in terms of certain works which - by their very great nature - encourage us to enter into relationship with them, seeking to know more of them, and it sometimes seems that they draw things out of us (in that an encounter with art can lead to self-discovery). Even though we have a kind of relationship with them, continued through study, or performances, each performance we attend may be likened to a conversation between the composer-in-the-work, about whom we are fascinated - because we love his work - and another person, who has something to say about them and to them. Thus, in experiencing a performance of a great work, we are witnessing a kind of dialogue, and if the performing speaker has a voice as significant and characterful as the composer, then the experience of audition may change our perceptions about them, and that about which they are talking. And sometimes, if the work really does point beyond itself, we may find that the ‘topic of conversation’ is me, or you, or the human condition, or God. When works point beyond themselves, and through the biographical or historical issues they draw upon, point to these issues; and when the performers are of sufficient stature; then we may, through recognition of some of the dimensions of the encounter, be changed, by what we subsequently call a truly great performance.

Summary

 

 

Having concluded my enquiry with a brief attempt to place it in a wider context, it is now appropriate to sum up all that we have seen and done. I introduced the idea of authentic performance, pointing out that while there is a wide range of literature on performance practice, often written by the musicians whose performances I have presumed to criticise, very little of it is of direct interest to my enquiry, since it does not engage on a philosophical level. I also gave a very brief historical perspective, in order to place the concept under discussion. I did not offer a definition of authentic performance, but rather characterised it in terms of the expectations which listeners and performers tend to have when encountering an ‘authentic performance’. Thus I narrowed down the focus of my enquiry to ‘authentic performance’, rather than ‘authenticity’, which we should not assume is the same thing, even though many musicians make this assumption.

Having selected the phenomenon for discussion I sought to defend it in Part One. As a starting point I took an article by Stan Godlovitch, in which he discusses some of the criticisms directed at authentic performers, and then makes a defensive reply, setting up a debate among three protagonists - the purists (whom he has been defending), the hodiernists and the liberals. The purists believe that old instruments and performance styles should be employed, and Godlovitch appeals to intention, functional fit and the essential nature of music to support them. The hodiernists believe that technological advances and the distance between us and the past renders authentic performance undesirable and impossible. He then mentions the liberals, who divide into two camps. The permissive liberal argues that period and modern performances are equally acceptable, hence that modern methods should not be sidelined by authentic performances. The missionary liberal (or missionary eclectic) actively pursues variety, affirming authentic performance as one of many approaches: great art is able to accommodate many interpretations.

In discussing Godlovitch’s paper, four assumptions were identified, on which much of his argument hinges. These assumptions do not figure as discussion points for him - either he is aware of them, and has made an informed decision; or he has unconsciously based his theories upon them. Either way, they are worth unpacking. The assumptions were:

1) that the work is identified with the score;

2) that the words written on the score as verbal instructions are as much part of the score as the notation (both of) which represent the composer’s intentions about how the piece should be performed.

3) The composer’s intentions are deemed to be relevant to performative interpretation.

4) Finally there is the assumption - even if these others are valid - that authentic performance is always to be preferred over other approaches.

Having outlined these assumptions, I discussed them at some length. Concerning the first assumption, the work of Nelson Goodman and Roman Ingarden was discussed, as representing two opposing views concerning the nature of the relationship between work and score. Goodman took the view that the score enables us to identify performances of a given work, since the notation has syntactic and semantic disjointness, syntactic and semantic differentiation, and does not accommodate ambiguity - each character has a unique determination. For a genuine instance of a work to be produced perfect compliance with the score must be achieved.

Goodman’s theory can be made practical by following Lydia Goehr’s modifications of it. Aware that a performance with a ‘mistake’ cannot be of the work under Goodman’s scheme, she proposes that a performance be of the work if it approximates perfect compliance and is intended to be recognisable as being of that work. Analytically this may be unsatisfactory, but it does at least make the performance of pieces possible.

Having seen how it may be possible to argue that the work can be identified in the score, I turned to Ingarden’s opposing views. For him, the score is intentionally designated with symbols which determine pitches, which themselves are not part of the score. The score is to be seen as little more than a set of instructions from composer to performer. In this Ingarden points to the use to which scores are most commonly put by performers. While we may wish to follow Goehr’s modification of Goodman on the whole, it is useful to acknowledge Ingarden’s point. Thus the score identifies a performance through the notion of intended compliance, but the very notion of compliance involves the use of that score as a set of instructions to performers.

Since Goehr’s modification allows us to work with an ideal for performance, we do not need to suppose that the supplementary, non-notational markings on the score are part of it, since Ingarden shows us that as instructions they carry intentional significance and relevance. Scores are not entirely notational - Diana Raffman highlights a weakness in Goodman’s theory by showing that the supplementary markings are semantically undifferentiated, which challenges his view that they are not part of the score in virtue of being semantically differentiated. This does not prove that they are part of the score, but it does re-open the possibility.

Questions about intention then come into focus. Having established that artworks are intentionally made, I turned to the classic work of Beardsley and Wimsatt, who claimed that the author’s intentions were unavailable and undesirable. Such a view presents a serious challenge to performers who believe that they are respecting the composer’s intentions through authentic performance. But Beardsley and Wimsatt’s theory succumbs to Colin Lyas’ account of ‘personal’ qualities, which presuppose an author, who intends the work to possess those qualities and could not do so had he or she not the power to do so. Thus the qualities demonstrate not only that the work has a creator, but also what kind of qualities the creator intended it to have.

Having suggested that the intentions of the author are available to us, I considered whether they are desirable or useful to us. From Mary Mothersill we learn that the question of relevance cannot be answered such that our interest in intentions must be disputed. Sometimes it serves no purpose, but at other times an interest in them can be beneficial. Furthermore, if the interest we have in works of art is similar to the interest we have in other people, then it is understandable that we should want to know as much about the object of interest as we can find out. We may ask ‘why did the artist do that?’, in which case we would be asking about the artist’s motivations, rather than merely discussing an aesthetic object divorced from its creator.

Much discussion about intentions focuses on criticism, not on performance. When we talk about interpretation we can be talking about criticism or performance. Jerold Levinson distinguishes between critical interpretations and performative interpretations, the latter of which is a considered way of playing the music. A PI can be repeated on different occasions and need not be instantiated in a single event. A purist looks for a singular best interpretation - the best CI instantiated in a single PI. If we adopt Goehr’s modifications we can acknowledge the existence of other attempts to approximate a performing ideal (other PIs), and we might recognise other CIs, although it would be possible not to do so in the opinion that there is an ideal performance to be striven for, towards which we are guided by the score and the correct CI. Finally, there is also scope for a multiplicity of PIs and CIs, although one PI cannot reflect the thinking of more than one CI at a time.

If we continue to identify the work with the score, then we acknowledge that the features of the score which express intentions - the verbal instructions - can be respected without going ‘outside the work’. Here again we may say that the intentions are available, but the question as to their desirability raises issues about the duties which a performer has to a composer and to an audience. A performance, claims J.O. Urmson takes place in a quasi-contractual situation, such that a performer is obliged to perform the piece stipulated - which means - according to Goehr - that he or she must intend to comply with the notation, and be recognised as doing so.

Thus, concerning this third assumption, there are good reasons for respecting the intentions of the composer - they can be valuable, and the arguments for not doing so are not conclusive. The other two assumptions were also acceptable - they can be defended - but when it comes to the fourth assumption the situation is different.