Musical formalism, which strictly limits the type of thing any description of the music can tell us, is ill-equipped to account for contemporary performance practice. If performative interpretations are in a position to tell us something about musical works—that is if performance is a kind of description, as Peter Kivy argues—then we have to loosen the restrictions on notions of musical relevance to make sense of performance. I argue that musical formalism, which strictly limits the type of thing any description of the music can tell us, is inconsistent with Kivy's quite compelling account of performance. This shows the difficulty that actual performances pose to overly rigid conceptions of music. Daniel Barenboim's unannounced performance of Wagner in Israel in 2001 shows that the problem of the boundaries of musical relevance is no mere philosophical puzzle. It is a pressing problem in the musical public sphere.
music, performance, musical work, performative interpretation, Peter Kivy, Daniel Barenboim, Wagner, Tristan and Isolde, Meistersinger, Israel, Nazis, formalism, public sphere, musical public sphere, politics, criticism, critical interpretation
In Music Alone Peter Kivy makes a peculiar, but compelling claim. He suggests that performative interpretations can give an account of music that, in some sense, bridges the divide between performative and critical interpretations noted by Richard Wollheim, Jerrold Levinson, and others. Kivy writes that "it is no paradox to say that performance is the ultimate nonverbal description of the work… I write of the listener's understanding; and that understanding, in the performer, is evinced most fully, most characteristically, in the nonverbal description that we call his or her interpretation—which is to say performance—of the musical work." That is, in some sense a performance of a work is the best way for us to come to know a work, to discover its various properties. I will argue that the formalism of Peter Kivy, which strictly limits the type of thing any description of the music can tell us, is inconsistent with his quite compelling account of performance. This inconsistency has implications that reach far beyond Kivy's exemplary account. It shows the difficulty that public performance poses to overly rigid conceptions of music. I will anchor my arguments in what I take to be a particularly rich example: Daniel Barenboim's recent performance of Wagner in Israel. This is the sort of hard case that clearly shows that the problem of the boundaries of musical relevance is no mere philosophical puzzle. It is very much alive as a problem in what I will call the musical public sphere.
2. Wagner in Israel
On July 7, 2001 Daniel Barenboim conducted the Prelude to Wagner's Tristan und Isolde as an unannounced encore at the Israel festival in Jerusalem. There has been an unofficial ban on, or what Barenboim prefers to call a "taboo" against, the public performance of Wagner's music in Israel. The ban had its origin in the Palestine Symphony Orchestra's cancellation of its performance of the Prelude to Die Meistersinger von Nürnberg in November of 1938. The cancellation was a direct response to the November progrom, which was euphemistically named "Kristallnacht" by Walter Funk, Chairman of the Nazi's Committee on Economic Policy, and Hitler's Personal Economic Advisor. The Palestine Symphony's refusal to perform Wagner and the general ban that followed served as a public condemnation of the racism of the Nazis.
In the years that followed, Wagner, as well as composers who were thought to have collaborated with the Nazis (Richard Strauss and Karl Orff in particular), were completely excluded from public concerts. While, interestingly, prime-time radio broadcasts helped slowly to lift the bans on public performances of Strauss and Orff in the 70s and 80s, the ban on Wagner has remained quite strong. It is well known that Wagner was the favorite composer of Hitler, and that Wagner's music was used prominently in propaganda events. Die Meistersinger was played at the Nürnberg rallies and countless other Nazi events, including several of Hitler's birthday celebrations in the thirties. Wagner's own nationalistic and anti-Semitic leanings are also well known from his essays and, though this is a more controversial claim, from the content of some of his operas.
In 1981, Zubin Mehta attempted to perform an excerpt from Tristan und Isolde unannounced and was stopped by the noisy intervention of the audience. For the 2001 Festival concert, Barenboim was originally scheduled to perform the first act of Die Walküre. However, protests by survivors of the camps as well as reservations expressed by the Israeli government led the festival authorities to ask Barenboim for an alternative program. Barenboim agreed to cancel Wagner and play Schumann's Fourth Symphony and Stravinsky's The Rite of Spring instead. The night of the performance Barenboim turned to the audience and told them that, as an encore, he would like to perform the Prelude of Wagner's Tristan und Isolde. He told the audience, as Mehta had, that anybody who did not want to hear it could leave. A 30 minute debate ensued where several people did leave, some noisily slamming doors and yelling "Juden raus!" the cry of the Nazi soldiers purging the ghettos of Europe. Nearly all of the remaining audience members gave Barenboim and the Berlin Staatskapelle a standing ovation after the encore.
The performance ignited a firestorm of debate. Conservative politicians responded with varying levels of vehemence. Then prime minister Ariel Sharon commented that he wished the work had not been played. Ehud Olmert, then mayor of Jerusalem, said that Barenboim's actions were "brazen, arrogant, uncivilized and insensitive" and that he had committed "cultural rape… It is not his job to determine whether the state of Israel decides to allow Wagner to be heard or not. As a musician he is great, but as a human being I could say a few other things." Zvulun Orlev the chairman of the committee on education and culture argued that Barenboim "should be declared a cultural 'persona non grata'"—that he should be banned from performing in Israel until he apologizes for his offense. Finally, Barenboim was officially rebuked by members of the Israeli parliament.
3. Performance, Criticism, and Philosophy
But does any of this tell us anything about how Barenboim actually performed Tristan? Does it tell us anything about his performative interpretation—his "account of the music"? I think it does, and I think nearly everybody who commented on the controversy, including Barenboim, would agree. To my knowledge, nobody has argued that all of this talk of politics is to miss the point of a performance, or to bring in outside considerations—extra-musical concerns—to criticize Barenboim's program choice. If one were to review the performance and not mention that it was Wagner being performed in Israel and instead merely give the usual descriptive and evaluative account of the event focusing on the quality of the playing, the shaping of particular phrases, the quality of tone color, and so on, it would either be a deeply inadequate review, not giving a full account of the performance, or it would be making a point. One might respond that, since all reviewing is interest-relative, I am here picking out one interest among many possible ones. It is possible that a hardened reviewer could have ignored everything but the sound structure of the work, or that another reviewer for a fashion magazine might have focused only on the outfits of the audience and performers. While it is true that interest is relative, it is my contention that any reviewer interested in the performance could not have ignored these factors. This seems uncontroversial. I am also claiming that any reviewer interested in the performative interpretation, a Wagner specialist interested in examining ways of interpreting Tristan, for example, could not ignore the political backdrop and the preconcert actions of Barenboim. This is, of course, more controversial. But I take it that any music reviewer of performances would be interested in these two things. Further, any music reviewer who had an interest in filtering out the political properties of the performance would either need to do so explicitly, and so be taking sides in the controversy and not really filtering it out, or would perform this argument with a silence that, in the context, could only be conspicuous.
The point that such a review might be making is that any properties of the event in question, unless they have something to do with the sounding of the music itself, are irrelevant to a description of the performance or a criticism of it as music. The music itself is, in the words of Eduard Hanslick, simply "tonally moving forms," where everything else is merely extra. One interested in making this argument could easily use the language of Peter Kivy's formalism that he so clearly sets out in Music Alone, or any number of articles in New Essays on Musical Understanding, or The Fine Art of Repetition. For example, in a "A New Music Criticism," Kivy joins musicologist Joseph Kerman in the call for a music criticism that moves beyond mere analysis, arguing that those who focus merely on technical description miss out on emotional properties that are essential to the work. Those who limit themselves to the purportedly dry technical language of music theory, though they wisely avoid the florid programmatic descriptions of the nineteenth century, end up throwing the emotional baby out with the extramusical bath water. However, Kivy emphasizes that it is crucial that the critic not go too far, sliding from proper criticism to what Kivy calls "interpretational criticism," which ought assiduously to be avoided. To interpret is to find meaning, and Kivy claims that in music there is no meaning to be found. Any description of music must be limited to calling attention to syntactical properties, and must not introduce a non-existent semantics.
Where does this leave performative interpretation? In Music Alone, Kivy writes,
[I]t is no paradox to say that performance is the ultimate nonverbal description of the work… I write of the listener's understanding; and that understanding, in the performer, is evinced most fully, most characteristically, in the nonverbal description that we call his or her interpretation—which is to say performance—of the musical work.
That is, in some sense a performance of a work is the best way for us to come to know a work, to discover its various properties. Presumably performances, as the "ultimate non-verbal description" of works of music, fall under the same formalist restriction with regard to semantics. Any interpretational performance must provide a false description of the work—it is a bad performance, or perhaps even a non-performance, if a performance is necessarily of the music. So, any performance would be inappropriate or in some sense false if it "ascribed" to the work semantic properties.
If ever there was a case of injection of extra-musical meaning into a work, arguing about the politics of the performance setting of a Wagner prelude would seem to be an example—the absurdity of a purely syntactic political predicate is manifest. We should be careful not to be misled by the presence of words in the actual opera. In fact, it seems that Barenboim chose his program carefully in order to make a justification for performing Wagner easier. He could have chosen the Prelude to Meistersinger, making his opponents' arguments easier and even, perhaps, agreeable to formalism. After all, the content of the words of Meistersinger is more easily argued to be politically offensive, dragging the music along with it. Tristan, though, is less obviously worrisome even in its lyrics and the Prelude is one of the most formally famous passages in all of music. I will say something about the significance of Barenboim's choice below.
It would seem that the formalists are compelled to insist on the same strictness with regard to semantics in the case of a performance of Wagner in Israel that they show toward descriptions of musical works in general. The political backdrop of the performance simply cannot be treated as musically relevant by a formalist because any properties of a performance for which it would be responsible would necessarily be semantic. That is, any properties of a performance event that are brought out by the political background of that event can not but be extramusical. It is important to set out this distinction between musical relevance and irrelevance with a bit more care. For an argument for the essential musical irrelevance of the political elements of a performance of Wagner in Israel to work, there must be certain properties that can make an aesthetic difference and certain properties that cannot. The argument assumes that there are types of properties of a performance that can be relevant and types of properties that cannot.
If it is true that interpretational performance is as illegitimate as interpretational criticism, it is important to have a clear picture of what this could mean. How could a formalist who held open the possibility of critical performances draw a line between an illegitimate interpretational performance-as-description (Barenboim conducting Wagner in Israel, Nazis conducting Wagner or Beethoven for political purposes, unificationists or nationalists in Germany using Beethoven's Ninth as an anthem) and a legitimate performance-as-description (a performance of the Ninth in Alice Tully Hall, Barenboim playing Tristan with the Chicago Symphony in Orchestra Hall). Of course, for normal verbal descriptions, critical interpretations in the language of Wollheim and Levinson, formalists would be happy to set strict, a priori conceptual constraints on the types of things one might say about the music itself. That is, formalists allow philosophy to do the conceptual work of dividing the world into the musical and the non-musical when it comes to writing about music.
4. Wigs and Politics
It is at this point where Kivy's own astute account of musical performance practice ought to cause a formalist account of music in general some difficulty. In a particularly rewarding passage of Authenticitie, Kivy considers what he dubs "the wig problem." The question arising from the wig problem is, How does one determine where aesthetically relevant gesture stops and aesthetically irrelevant content begins in a performance? Kivy gives an account of aesthetically relevant, musically relevant—he sensibly collapses the two when considering appreciation of performance—visual aspects of performance. He argues that certain visual properties of performances are indeed musically relevant. How are we to distinguish between musical gestures and extra-musical content of performances? He suggests that it is possible in any number of cases to argue, to give a viable interpretation of the music that makes extra-score or non-sonic properties musically relevant. But, he concedes, this relevance is entirely dependent on particular arguments and local justifications. Kivy steadfastly refuses to set out criteria of musical relevance with regard to performance—it is simply not appropriate definitively to set parameters in advance of actual performances. If one can make a reasonable case for the aesthetic relevance of some performance choice, he argues, then that is all there is to it.
It would be helpful to consider in some detail a few of Kivy's own illuminating examples. At the end of Haydn's Symphony Number 45, "The Farewell," the musicians stop playing one by one, blowing out the candles on the stand and leaving with their instruments under their arms as they do. Kivy gives a compelling argument as to why the movement is relevant, and why it should be done with candles instead of electric music stand lights. He argues that blowing out a candle is a more fluid, graceful movement than snapping off a stand light and this motion is far more suited to the quiet adagio playing. I would add that the dramatic effect of extinguishing a flame with one's breath is far greater than snapping off an electric light, no matter how smoothly and gracefully one might turn off the light. The rising smoke dissipating into the air accompanies the player as she exits the stage, leaving behind a rapidly fading memory. Kivy also suggests that he would accept that a performance of Mozart's Coronation Mass in a "candlelit Salzburg Cathedral" is more "musically authentic" than performing it in the sonic museum of a contemporary concert hall. This is not to say one should not or can not perform the work in the sonic museum—there is plenty of pure music that the work has to give that would allow it to withstand being played in a modern concert setting. However, playing it in a more traditional setting like the Salzburg Cathedral brings out musical properties that would not otherwise be apparent. "I can give a plausible story about how the structure of the Coronation Mass and its shining musical surface have been aesthetically fashioned for ceremonial performance in a place of worship of a certain kind, " says Kivy.
In each of these examples, elements of visual appearance, the performers' actions, and the performance context contribute to the understanding of the music itself. On the one hand, for a performance to be a legitimate, non-interpretational description on Kivy's account of musical meaning, the content of this description must be merely syntactical. As such, it can not amount to very much in the grand scheme of things. Kivy does not mean this to denigrate the performer or performance, just as arguing that there is nothing profound in music is not to denigrate music—it is just to have a good conceptual grasp of the kind of thing performance or music is. Profundity is reserved for things with semantic content, with things capable of telling us something new and deep. As Kivy has repeatedly insisted, music on its own falls short on all these counts. On the other hand, coupled with the examples from Authenticities, the statement that performance is a description of the work pulls us in the opposite and, I would argue, in a more appealing direction. In the examples helping to illuminate the wig problem, it is not only admitted that a number of visual, gestural, and contextual elements of the performance tell us immediately and viscerally about the work; they can also expand our understanding of the work. A wide variety of properties of the performance—sonic or otherwise—can aid us in refining the identification and evaluations of the properties of the work itself. That is, properties of many different types can be musically relevant in a performance. The "shining musical surface" of the Coronation Mass looks and sounds quite different in a Salzburg Cathedral and one might learn something about the work that one could not have otherwise. One might feel more justified in describing its solemn grandeur as "opulent," where one may not have before, perhaps settling for "complex," or "intricate," "densely interwoven," etc. The justification for the description would in part depend on the performance's aesthetically working in a particular performance setting along with the usual technical analytical points about the score. The spareness of the second movement of the Farewell Symphony becomes much more poignant and pointed at the end of the last movement. The musicians getting up and leaving after blowing out the candles while the music still plays adds another dimension to the emotional content of the performance that would simply not be there without the actions.
It might be tempting to argue that I have simply missed the point by missing a distinction. While the distinction between something that can be musically relevant and something that cannot be is a matter of the concept of music, the distinction between musical relevance and irrelevance of performance choices assumes the properties at stake have passed conceptual muster. That is, among the properties that can be relevant, some are in certain circumstances and some are not. For example, the performer's wearing wigs can be an aesthetic property—one can imagine a situation in which it makes an aesthetic difference whether or not wigs are worn. Similarly, one can imagine situations where blowing out a candle on the stand is utterly irrelevant. Similar things might be said about more general scene setting. Whether a work is performed in a cathedral clearly can be aesthetically relevant. In the particular case of the Coronation Mass, it is relevant. In the case of Schumann's Fourth Symphony, it is not. The religious cum political pomp of the Coronation Mass becomes more prominent and takes on a particular hue in a cathedral. Imagine it instead in a spare, though acoustically identical German Lutheran church—the musical surface still shines, but differently: without the pomp and opulence—perhaps it is even ironic, or mocking.
In the case of Barenboim's performance, then, one can take three positions. One might argue that (1) the political surroundings cannot make an aesthetic difference, that is, they cannot be aesthetic properties at all, or (2) they can be aesthetic properties but are not in this case, or (3) they can be and are. The argument I began with, in the spirit of Kivy's formalism, depended on (1). When Kivy writes of musical works themselves, he argues that political and moral matters are necessarily matters of semantic content that music alone cannot have. Thus they cannot be aesthetic properties of music, since these properties comprise only formal and syntactic ones. But Kivy's own arguments give us powerful reasons for doubting the wisdom of making such categorical judgments in advance of actual performances. That is, the "plausible story" criterion of musical relevance in principle leaves open precisely what properties can be musically relevant. Since it is impossible to rule anything out, it is possible for anything to be musically relevant. His arguments about musically relevant performance choices suggest that the real argument, when we discuss performances, is between (2) and (3). Of course, certain types of properties may so seldom present themselves as candidates for musical relevance as to seem in principle irrelevant. If this is in principle true of performances of any work, and performances are descriptions of works—that is, they pick out aesthetically relevant properties of works, then it is impossible to see how the properties of the work itself can avoid being subject to the same criterion of relevance. It seems, then, that Kivy's compelling justifications of musically relevant properties of performances entangle the content of the performed works in a widening net of semantically fraught context.
In performances, the bright line between essentially irrelevant properties and properties that can be relevant dims considerably. In fact, it seems that there is no principled way of determining just what will be or can be aesthetically relevant to a given performance. Recall that whether the Coronation Mass is better performed in a cathedral depends on an interpretation of the music that was made for "ceremonial performance in a place of worship of a certain kind." Note how specific Kivy allows the musical properties to be here: Ceremonial performance (as opposed to merely formal, say) in a place of worship (as opposed to any old hall amenable to a certain level of solemnity) of a particular kind (what kind? Catholic? opulent?). Kivy is absolutely correct in arguing that such details are often unavoidable when considering aesthetically relevant performance choice. Whatever the details of these descriptions, it is impossible to characterize all of these ever more specific properties as merely syntactical. Consideration of these details leaves open questions that it would seem formalism would in other contexts like to close. More important, if performance is to have the artistic role that it does for Kivy, these elements cannot be determined in advance. If a place of worship of a particular kind can count as an aesthetically relevant factor of a musical performance, again, I see no principled reason for rejecting a priori that a particular political context could also be. How fine grained this political context might be seems to be open to argument by performers and their publics.
5. Generative Context
One might respond that any content open to discussion as musical content, content of the work itself, in a performance must be plausibly connected to the generative context of the work. In the case of Tristan, that the work is performed by Barenboim in Israel, and that this is politically significant, tells us nothing about the work—though perhaps it tells us something about Nazis, Barenboim, and Israel. The reason that the work is left out of this account is because the performance choice, or performance context, in this instance is different in kind from that of the Coronation Mass or the Farewell Symphony. In the latter two cases, the Salzburg Cathedral and the blowing out of candles are directly connected to the intentions of the composer and the original context of the work. This is implied by Kivy, whose plausible story is about how the work is fashioned in a particular way. The political context of Israel, one might think, has nothing to do with the generative context of Tristan. And so, on this account, there is a principled distinction to be made for musical relevance, and it is this: only something that can be plausibly connected to the generative context of the work itself can be musically relevant.
Unfortunately, this response simply begs the question. Barenboim's performance in Israel is controversial precisely because it there is no uncontroversially pre-given account of generative context of the work. In fact, it is controversial precisely because it makes a sort of claim about it. Moreover, I want to argue that this could be said of any performance, even though I readily admit that the vast majority of performances are uncontroversial and the generative context of the work is not explicitly addressed as a problem. In this sense, most performances are like the all-too-common "easy cases" in law that can make legal decisions appear to be automatic and amenable to formalistic interpretation. Sophisticated arguments against the performance of Wagner in Israel argue that the music itself contains the German nationalistic, anti-Semitic content that makes the music in itself politically and morally repellent. This claim may well be (and almost always is) justified at least in part by the generative context of the works in question. Reference to Wagner's avowed anti-Semitism, the culture of anti-Semitism in which he wrote, the explicit nationalism in his writings and perhaps even in works like Meistersinger, are part of what allow those against the performance of Wagner to identify his work in the way that they do. That these parts of the generative context have been emphasized (by Nazis, Israelis, scholars, journalists, and so on) is what is called into question by Barenboim. His writings as well as his performance of Tristan argue that this is not how the work should be understood, that these undeniable elements of the generative context of the work are not relevant to the identity of the work. On Barenboim's account of the work, and the generative context relevant to the identification of its salient properties, those who argue otherwise have simply been making a mistake. The argument for and against a political understanding make the same use of generative context, but they call attention to or emphasize the relevance of different parts of it.
If we take seriously Kivy's account of the performance as description of the work, and we add that the work is in part determined by its generative context, then the performance, insofar as it successfully describes the work, commits itself to a particular understanding of the work's generative context. This seems odd only on a wrong-headed view about what the generative context is, what can be settled about it, and how. A full account of what I have in mind here would take me beyond the scope of this paper, but I believe I can give an indication of the argument I have in mind.
What it is about that context that is relevant to the identification of the work is not a simple matter of historical explanation. In hard cases, how we should appropriately understand history may well be part of the dispute, rather than the neutral ground on which decisions of either side can be made. Nazi performers, insofar as they were coherent, were trying to say something about the work itself. And, of course, they believed that this would also tell us something about German character, and a great many other things. Barenboim's performance of Tristan in Israel was a powerful way of denying that Wagner's anti-Semitism and nationalism were relevant to, or perhaps just determinative of, the identity of the work. I am happy to admit that some connection to the generative context might be crucially important, or even dispositive in certain cases, in our investigation of musically relevant performance properties. But this concession does no harm to my position. If the performance is a description of the work, an essential aid to our coming to an understanding of the work, it participates in this very determination. So the generative context cannot be used as a definitive conceptual arbiter between disputants over musical relevance in a performance.
6. Barenboim, Tristan, and Meistersinger
What, then, did Barenboim's performance of Tristan in Israel tell his audience something about Tristan that contributed to their musical understanding? Among a great many other things, Barenboim's performance described Tristan und Isolde as not itself politically repugnant. When asked in a public lecture recently whether he thought that Tristan had any anti-Semitic content, Barenboim replied, in effect, "Of course not. I would not perform it if I thought it did." This may seem quite weak as it stands, but given the continuing controversy over Wagner in contemporary musicology and the history of his music's reception in Israel, even this apparently weak, negative claim ranges wider than one would expect. Of course, it does seem simply to tell us what the formalist insists was clear all along. But Barenboim's performance is doing something more and different from the formalist. His argument functions as an instance of an argument of form (2) above—political content can be musically relevant, but a it is not in this case. If he were just making the general formalist claim that there could be no musically relevant political content, then Meistersinger would have done the job equally well, and more shockingly. But he chose Tristan.
Since one of the relevant facts of performance that helps us understand a performative interpretation is program choice, it may be helpful to end by considering why Barenboim might have made this particular choice, and what this might tell us about Tristan. What a performance of Tristan in Israel highlights better than a performance of it elsewhere is the nature of the struggle of the music—it lends it an inflection it might not have had otherwise. In Tristan, the desire and longing of the lovers bubbles up as a beautiful and compelling disorder against the externally imposed social order of the world. The same is true of the individuality, talent, and artistic genius of Walther in Meistersinger, and the purity of Siegfried in the Ring cycle. The order of the world stands as a constraining obstacle to the expression and satisfaction of endless desire, yearning, longing, talent, true art, and greatness. The character of each of these obstacles is quite different in each work, of course, as is the nature of the struggle. Think what happens in the music of Meistersinger. The music shows a struggle between free-flowing, easy breathing, vibrant, self-generating, natural melodies and oppressive, stodgy, archaic, foreign, imposed and unnatural rules of harmony. After the Prelude, David, the star pupil of the pedantic wing of the Meistersinger guild, enters with a long, painful, and tedious aria. Its use of modality sounds archaic (though it isn't), stodgy, and foreign to the listener. It is finally interrupted with the opening aria of Walther. Walther's music, of course, is like a breath of fresh air—gone are the modes, gone is the stodginess, and in comes free-breathing, natural-sounding singing. As the opera progresses, of course, the rules and strictures imposed by the guild are shown to be ridiculous (though we already heard that in the opening arias) and they are overcome and defeated by Walther who is ultimately praised to no end by the unified voice of the community, minus Beckmesser of course. The triumphal univocity of the great C major chorus at the end of Meistersinger signals overcoming the stifling old order, and the exclusion of Beckmesser's pedantry, through the emergence and expression of a people's natural voice—this voice of the crowd has been there from the beginning, but it brought to its full unity only by Walther's natural genius.
This strident, triumphal, univocity of the crowd (distinct from Walther's natural melodies) and the exclusion of the archaic Beckmesser is precisely what was highlighted and given specific inflection by performance in a Nazi setting. In such a setting, we see the triumphal univocity as the identification of a mass, instead of a public. In fact, it is difficult and ought to be difficult to see the crowd in Meistersinger as anything but a mass. This is not to argue that identification with the mass, or sympathy with it, is the point of the final scene, let alone the entire opera—I do not think that the Nazi's were correct to reduce Meistersinger to this. But it seems to me that the disturbing element is clearly there in the stridency, triumphalism, and univocity, achieved only through exclusion, of the music itself. What is made of this element depends on the plausible story one tells, either through critical interpretation or through performance, about the music itself.
The overcoming of opposition in Tristan is of course strikingly different. There is no sense of mass in Tristan and Isolde's "triumph" over the order of the world. Though an overwhelming desire and longing are present and persist throughout, they are brought out with constant musical and emotional shifting. The final resolution doesn't overcome the world and it doesn't incite a crowd to adoration, self affirmation, or mass triumphalism. It makes the world irrelevant through the transfiguration of Tristan's and Isolde's longing in death. The work is staunchly opposed to the trouble-free synthesis of individuals and society, or even a trouble-free synthesis of a conflicted individual. Though one might argue that the same can be said of Meistersinger, much more argumentative work would need to be done to overcome the apparently strident syntheses. That is, it is much easier to generate a plausible story whose elements are uncontroversially non-political in the case of Tristan than it is in the case of Meistersinger.
Description and evaluation of performances provide us with a particularly clear view of some of the shortcomings of formalist accounts of music. Kivy's insight into the role of performances in our coming to understand music, combined with his rich descriptions of performances of works, serves to undermine his own formalism. The audience at the Israel Festival understood what Barenboim described in his performance and contested his account. Nobody suggested that as "music alone" it had nothing to do with "the world of the world." Barenboim argues that we ought to listen to Wagner's music alone, and that it has something else to do with the world of the world than many people in Israel and around the world think it does. What is fascinating about Barenboim's performance is that the political context is directly relevant to his musical point that the music itself does not contain Wagner's repugnant politics. Had the performance just assumed the general formalist point, Barenboim, or anybody else, would have no reason to call such attention to the fact that he was performing Wagner in Israel. The commitment to listen to the music itself, and give oneself over to the world of "infinite longing" opened by Wagner's Tristan, asks audience members to put aside their "terrible associations." It asks them to commit to entering into the musical world and leave behind, for the time being, experiences that most Israelis would consider constitutive of their identities. Barenboim and his audience were perfectly correct to see his performance as a politically charged though nevertheless a musical gesture.
It is relatively uncontroversial that part of the role of performance is to challenge our understanding of a work. What I have suggested here is that this challenge not infrequently goes both wider and deeper than formalism can countenance. Musical performance often presents us with hard cases that force us to question what is musical, what is part of a work, what role music can play for us, what role music does play for us, what commitments I must have to listen the way I do. Barenboim's performance of Wagner in Jerusalem directly raised all of these questions and offered answers to some of them. It is not possible to make sense of the significance of hard cases in performance practice without leaving open the sorts of things that might be musically relevant. In Israel, Barenboim made vivid the depth and breadth of a commitment to hearing music alone in a way that could not have been done elsewhere. His performance of Wagner involved political, moral, and historical commitments all of which inform and reveal a conception of the work itself. Barenboim made a powerful statement about Wagner, or at least about Tristan, that could not have been made elsewhere. Whether his story about Wagner and Israel is a plausible one is still an open question to be settled only in the ongoing deliberation in the sphere of the musical public.
Jonathan A. Neufeld email@example.com
Jonathan A. Neufeld is Assistant Professor of Philosophy Vanderbilt University, Nashville, TN 37240.
Published February 27, 2009
 Music Alone, (Ithaca: Cornell University Press, 1991), p. 122.
 My account is heavily indebted to Na'ama Sheffi's fascinating and detailed history of the "Wagner problem" in Israeli culture and politics, The Ring of Myths (Brighton: Sussex Academic Press, 2001). Whether or not Strauss and Orff collaborated with the Nazis is irrelevant to my claims here. For an interesting account of music in Germany in the Nazi era, see Michael Kater's Composers of the Nazi Era: Eight Portraits (New York: Oxford University Press, 2000) and The Twisted Muse: Musicians and their Music in the Third Reich (New York: Oxford University Press 1997).
 See "Jews in Music," in Goldman and Sprinchorn, eds., Wagner on Music and Drama (New York: Da Capo Press, 1964), pp. 51-59. Die Meistersinger von Nürnberg is the most obvious and least controversial (though not uncontroversial) example of an operas containing anti-Semitic content in the characterization of the unmusical pedant Beckmesser and his star pupil David. See also Marc Weiner, Wagner and the Anti-Semitic Imagination (Lincoln: University of Nebraska Press, 1995), for a detailed argument that anti-Semitism permeates not only the libretti of Wagner's operas, but much of Wagner's music considered more generally.
 See Ewen MacAskill, "Barenboim stirs up Israeli storm by playing Wagner" in Guardian, Monday, July 9, 2001, archived in Guardian Unlimited online: http://www.guardian.co.uk/Archive/Article/0,4273,4218104,00.html; accessed 11/21/2006.
 I thank Peter Lamarque for suggesting I address these other interests a reviewer might have.
 See Lydia Goehr's The Quest for Voice (Oxford: Oxford University Press, 1998) for an analysis of the concept of the "extra-musical."
 Fine Art of Repetition (Cambridge: Cambridge University Press, 1993), pp. 296-326.
 Among which general emotional properties are included for Kivy, setting his account apart from other formalists such as Eduard Hanslick.
 Music Alone, p. 122.
 Although Jerrold Levinson has pointed out that "nonverbal description" may be self contradictory, I would like to leave this worry aside for the moment and simply accept that the thought that a performance can tell us something about a work is at least a difficult intuition to deny.
 Authenticities, p. 102.
 Kivy argues this in Music Alone, Chapter 8, Fine Art of Repetition, Chapter 13.
 Kivy also seems skeptical that any artwork, including literature, can be profound.
 I am grateful to Brian Soucek for raising this objection.
 Not every argument that acknowledges the political content of Wagner needs to argue against its performance in Israel, of course.
 See Weiner, Wagner and the Anti-Semitic Imagination, and Adorno, In Search of Wagner (London Verso, 1981).
 The exchange was relayed to me by Lydia Goehr, who was also the questioner.
 See Lydia Goehr's The Quest for Voice (Oxford: OUP 1996).
 Kivy, Philosophies of Arts (Cambridge: Cambridge University Press, 1997), p. 207.
 A version of this was delivered at the American Society for Aesthetics, Pacific Division. Thanks to Tobyn DeMarco, who was my commentator. Thanks also to Lydia Goehr, Brian Soucek, Hanne Appelqvist, Michalle Gal, Sirine Shebaya, Tiger Roholt, and Gregg Horowitz for their extensive feedback and support.