Does Performance Matter?
It is a commonplace idea that performance is of great importance in the communication of serious music. In a sense, it is easy to see why this should be so; for most people, music can only be heard when it is performed. But the value attributed nowadays to performance goes beyond this practical consideration. Indeed, there can be little doubt that we live in a musical age which exalts the performer, renders him fame and fortune far in excess of what is granted to living composers, and even confers upon a lucky few a measure of immortality through the medium of recording.
As a pianist, I have spent much of my own musical life in an atmosphere carefully structured to produce active believers in this cult of performance. I was trained mostly by Russian teachers, ending with three years of lessons from the queen of Juilliard piano teachers, Rosina Lhevinne, who died in 1976 at the age of ninety-six. From her and from my previous teachers I received a rigorously strict education in the way the piano was played in Russia even before the turn of the century; viewers of the recent films on the Kirov ballet school and on Russian gymnasts will have some idea of what this training was like in spirit even if not in specific content.
More was involved than simply learning how to master the piano mechanically. What to do with every piece—which of the innumerable technical and musical options to employ—was a matter of equal importance. How seriously such details were taken for their effect on the final result may be seen from a tense exchange I had with Mme. Lhevinne during one of my lessons in the early 1960’s. I was playing for her the Emperor Concerto of Beethoven, a work which I was to perform the following week with a minor orchestra in California. As I started to play the opening cadenza, she immediately objected that I was using the damper pedal—the mechanism on the piano, operated with the right foot, which sustains the sounds after the fingers have released the keys—much too much. When I demurred, she said: “But last night you went to hear Richter: didn’t you hear how he used the pedal?” When I answered that the famous Soviet pianist’s pedaling was the one aspect of his playing I disliked, Mme. Lhevinne softly answered, after a pregnant and hostile pause: “I only wish for you the same success.”
Every detail, then, was of major importance to the performance as a whole. I myself accepted this general proposition in evaluating both my own work and the work of others. And so, it seemed, did everyone else. Thus, whenever I went before the public I found, as does every performer, that I was critically judged in terms of how well I had satisfied the finely detailed standards of each reviewer, standards which were based upon the most careful discrimination of difference. The problem was that these differences themselves differed very widely from reviewer to reviewer.
There was one school of performance which tried to solve the problem of difference with a single intellectual stroke. This was the school of historical authenticity, preeminently associated among pianists with the name of Artur Schnabel. To Schnabel and his followers, the right way to perform a piece of music was not the way decided by the performer (or by his teachers!); the right way—and the only way—had been decided by the composer in the very process of writing his piece. For proponents of this position, there was no longer any discussion of Rubinstein’s tone versus Horowitz’s, of Hofmann’s staccato versus Rosenthal’s, or even Rachmaninoff’s Chopin versus Cortot’s. For these fundamentalists, the question was not how an artist expressed himself through his own interpretation, but whether he expressed the music through fidelity to the composer’s intention.
In practical terms, this meant playing fully, no more and no less, what was notated in the best possible edition of the music, and the subordination of the performer to the music. Yet even this attitude of fidelity to the text (called texttreue in German) turned out to be only another means of identifying differences and assigning importance to them. That these differences were seen as involving authenticity rather than virtuosity and self-expression hardly meant that they no longer counted: indeed, from this point of view, the music depended even more heavily on performance (the “right” performance) than ever before.
For a long time now I have been nagged by doubts about the cult of performance, but it is only since I began listening to many more performances than I once had occasion to do, and to listen to them, as a critic, in quite a new way, that these doubts have begun to crystallize. Recently, for example, in preparation for an article on Wagner’s Ring,1 I listened with score at least once to nine different complete performances of the entire cycle—approximately 135 hours of music—as well as to most of the recorded performances of one or another part. As I listened, the music remained beautiful, but to my surprise the specific performances seemed to count less. Except for some miserable singing on one set of recordings, and the intrusion of English on another, all the versions were more than adequate to allow for the deepest musical pleasure; but beyond that, the highly talented individuals involved tended to become increasingly vague in my mind as compared with the power of the music itself. The differences in performance—here faster, there slower; here louder, there softer; here a singer with a better voice, there a singer with a more passionate characterization—were essentially nugatory.
It was the same with recording quality and technique. The relative range of quality in those recordings was wide, an understandable product of their varied ages and the original purpose for which they were made. But a very few minutes of listening to each recording was enough to accustom me to its technological level. Here as elsewhere, as my catalogue of differences lengthened, my assessment of their importance became weaker. The music still came through, and the music was what mattered the most.
But this was opera. What about other kinds of music? Would the same conclusion emerge from a similarly concentrated exposure to many different performances of, say, symphonic works? In an effort to answer the question, I chose for consideration perhaps the most widely played instrumental music—and some of the most beautiful and influential as well—ever written: the nine Beethoven symphonies.
The current Schwann catalogues list 18 versions of the complete set; more are available in better record stores from European companies which have not chosen to have them widely distributed here, and still others have been made and are now discontinued, awaiting repackaging and reissuing at lower prices. In addition to all this, there are countless recordings of individual symphonies by every major conductor and many others little known or even ignored.
Out of this embarrassment of riches I have listened to ten complete sets: one each by (in alphabetical order) Ernest Ansermet, Bernard Haitink, Otto Klemperer, Erich Leinsdorf, Willem Mengelberg, George Szell, Arturo Toscanini, and Bruno Walter—and two by Herbert von Karajan.2 I have also listened to individual symphonies performed, sometimes more than once, by Pierre Boulez, Colin Davis, Wilhelm Furtwängler, Jascha Horenstein, Karajan, Erich Kleiber, Klemperer, Pierre Monteux, Arthur Nikisch, Fritz Reiner, Victor De Sabata, Hermann Scherchen, Leopold Stokowski, Toscanini, Walter, and Felix Weingartner.
All these recordings are well known, and some of them are famous. They have been exhaustively and repeatedly described by highly competent critics both in contemporary reviews and later in fully annotated articles and discographies. Though there is always some, and occasionally a great deal of, disagreement among these evaluations, a consensus about each conductor’s work tends to emerge with which it would be difficult to quarrel. Toscanini’s records are fast, bright, tense, and often exciting; Mengelberg’s are wayward, arbitrary, distorted, and provocative; Klemperer’s are serious, humorless, heavy, and “profound”; Szell is vital, interesting, somewhat plain, and concerned with getting the best possible playing from his orchestra; Karajan is brilliant, a bit slick, and strangely moving in his icy control. And so it goes.
Among the performances of single symphonies, Boulez3 in the Fifth sounds remarkably uncommitted either to the music or to the tradition which has grown around it; Furtwängler4 unconsciously tears the music apart and therefore sometimes achieves effects of the utmost plasticity; Scherchen5 takes some unusually fast tempi in the Eroica; Weingartner6 offers a curious combination of unmannered conceptions with lush orchestral executions. And so on.
The differences, then, are certainly there in the large. And they are also present in detail, as I discovered by selecting two passages—one in the First Symphony and the other in the Seventh—which pose two distinct kinds of problems in performance: one in the conductor’s fidelity to the text, and the other in the level of orchestral playing which he is able to elicit.
The passage in the First Symphony is the second theme of the first movement. Here the problem for the conductor is to find a tempo which will both accommodate the tranquil opening of the theme and its weightier and more outspoken continuation, and at the same time not be so different from-the pace of the rather skittish first theme as to lose the possibility of overall integration.
Of the recordings I have listened to of this section—14 in all—the greatest contrast is between Mengelberg and Leinsdorf. Mengelberg, though he does not take a slower tempo to begin the second theme, adds a ritenuto—an immediate slowing down—to the bars introducing the weightier music that follows. Not content with that slackening, he concludes the phrase with an enormous ritardando—a gradual slowing—on the fortissimo chords which introduce the return to either the semi-obligatory repeat or, after the repeat has been played, to the development section of the movement. But Leinsdorf, possibly the most textually observant of the conductors I heard here, will have none of this alteration of tempo. Since the score indicates no change in this particular section, Leinsdorf plays it in the very fast Allegro con brio tempo which Beethoven prescribes for the main body of the movement.
Of the other conductors, both Klemperer and Haitink are as rigid as Leinsdorf, but in a slower overall tempo. Haitink goes along at a middle-of-the-road pace while Klemperer chooses a tempo so slow as to give a ponderous and dour impression. Szell, Monteux,7 and the two Toscanini versions8 are, in ascending order, slightly more flexible rhythmically, though always within a strict classical framework. Karajan in both his later recordings shows a greater willingness to take liberties, but these are pale compared to Mengelberg’s. Walter and Furtwängler, though they bear some resemblance to Mengelberg in their perception of the demands of the music, are carefully discreet in their deviations from a single tempo.
Do these differences in performance make any difference to the experience of listening to the music? I cannot see that they do to any significant extent. Even Mengelberg, though his recordings are currently the subject of attacks in the musical press for what is considered their interpretative license, hardly seems to be doing more than underlining what Beethoven wrote.
The opening of the last movement of the Seventh Symphony presents difficulties not so much of musical interpretation as of execution. Chief among these at the start are the articulation of winds and brasses in the repeated notes and the clarity of the first violins in several short scalic patterns. Coming slightly later are the fast dotted rhythms for the strings and the rapid two-note phrases in the first violins. In both passages, not only clarity but good tone is the goal.
Of the 15 versions I listened to, the Cleveland Orchestra under Szell undoubtedly surmounts the problems of these passages better than any other group; the Orchestre de la Suisse Romande under Ansermet is the poorest performer. Both in clarity and tone quality alike, all the other orchestras—including such legendary groups as the Boston Symphony (Leinsdorf), the Concertgebouw of Amsterdam (Mengelberg), the 1937 New York Philharmonic (Toscanini),9 the NBC Symphony (Toscanini), and the Berlin (Karajan) and Vienna (Furtwängler and Karajan) Philharmonics—fall in between.
Nevertheless, despite these perceptible differences, all in all the better performances are only marginally clearer, cleaner, and of more pleasing sound than the lesser efforts. And in any case, no orchestra, including the Cleveland, plays the passage perfectly, and no orchestra, including the Suisse Romande, plays it so badly as to interfere to any serious degree with the music. All the performances I heard adequately communicate the structure of the music and its emotional power.
Of course it may be that this is so only because the average level of artistic achievement on these records is very high. Yet there are grounds for thinking that even differences vastly greater than those which are present in these Beethoven recordings matter only marginally. Two kinds of evidence, admittedly speculative, may be adduced here. The first concerns what may be reasonably suspected about the early performances of what we now accept as the great master-works, and the second involves the curious phenomenon of transcriptions.
Anyone who has had anything to do with the performance of new music knows just how tacky most of these renditions actually are, and how approximate is the fulfillment of the composer’s notes and performance directions. I wish I could say that I am talking here only about subtle matters, but in fact veterans of such performances know in their hearts that survival often must replace fidelity as a goal in this kind of work. While modern music is indeed hard to perform, there is no reason to suppose that it is harder for us than the new music of the 19th century was for the performers who first had to play it. Certainly composers, then as now, complained bitterly about the performances they were accorded. It would therefore seem likely that such works as the Beethoven symphonies were themselves mangled on the concert stage early in their lives, subject as they must have been to lassitude, perversion, and simple incompetence. And yet the masterpieces established themselves—to the point that present-day performers are often accused of lèse majesté if they fail to play them to the highest possible standards.
The scandal about transcriptions has now mostly died down, not because they have been accepted but rather because they have been driven out of existence in polite society. So completely have they vanished from orchestra programs and the repertories of refined instrumentalists, that it is easy to forget how much the most famous music in the Western tradition owes to its being thus retailed. In the 19th century, in the absence of recordings and of today’s many concerts, orchestral and chamber compositions were circulated in transcription designed for amateur four-hand piano performance in the home. Professional soloists also made a practice of playing transcriptions of operatic and even orchestral excerpts. All this helped to keep the music alive. In more recent times, Bach’s popularity owes much to the inflated and often bombastic transcriptions made by and for Leopold Stokowski and other famous conductors of modern orchestras. (Something of this practice continues to exist, though outside the world of serious music: electronic, “synthesized” performances of both Bach and Debussy have become best-sellers in the pop market.) In short, the music has survived, and indeed prevailed over, even the most vulgar distortions of performance.
Another example of the power of great music to survive inadequate performance may be seen in the remarkable recording of the Beethoven Fifth made by Arthur Nikisch10 with the Berlin Philharmonic in 1913. In this performance, because of the inability of the primitive acoustic recording process to capture sufficient sound from the strings, brass and wind instruments were used to add volume. Yet what emerges from the record is the symphony itself. It is not the work played by an orchestra in a live concert, not a modern recording, but it is, simply and powerfully, the very core of what Beethoven created.
No one, surely, would argue for the positive virtue of stone-age recording processes or incompetent performances. But recording techniques have long since ceased to be a problem, and even competence is becoming less and less of an issue. There was a time when a performer could either get through a piece—playing as many of the notes as correctly as possible, as audibly as possible, and as musically as possible—or he could not. Records have changed all that. Since the late 1940’s, magnetic tape has enabled each note, if necessary, to be recorded separately and then, thanks to the miracle of inaudible splices, assembled into a “whole” performance. When this convenience is added to the electronic manipulation of sound level, tone color, and room ambience, the very possibility of incompetence in the old sense disappears. The weak sound strong, the old young, the dull bright, and the sloppy precise.
Thus, on records, technically excellent performances have become as routine as they are—and always have been—rare in the concert hall. Perhaps it is this above all which has made performance, in spite of its present position in musical life, seem so much less consequential than once upon a time it did.
1 “Performing the Ring,” COMMENTARY, January 1978.
2 Ansermet: London STS-BTH-S-1; Haitink: Philips 6747307; Klemperer: Angel SH-3619; Leinsdorf: RCA VCS-6903; Mengelberg: Philips 6767003; Szell: Columbia M7X-30281; Toscanini: RCA VIC-8000; Walter: Odyssey Y7-30051. The earlier of the two Karajan recordings is listed as Deutsche Grammophon 2721001, and his just released newer set is Deutsche Grammophon 2740172.
3 Columbia M-30085
4 Many Furtwängler performances of Beethoven symphonies may presently be purchased on both commercial and private recordings. In this country, Seraphim has released the Third (Eroica), Fifth, and Seventh Symphonies in album 1C-6018; the Ninth is available, also from Seraphim, in album 1B-6068. Still freer performances are irregularly available from Olympic Records in this country, and on the Deutsche Grammophon and Unicorn labels abroad.
5 Westminster WG-8352.
6 The Eighth and Ninth Symphonies are sold in this country on Turnabout THS 65076/77; the Third is currently available only in Germany, on Electrola 1C 053-01 481.
7 London STS-15238.
8 The earlier Toscanini version is Seraphim 1C-6015; the later performance is that of the complete set listed above.
9 RCA VIC-1502.
10 Elcctrola 1C 053-01 466.