Andrew Porter, The New Yorker, October 24, 1983
When a work of Beethoven's was performed, his first question was always "How were the tempi?", according to his biographer Anton Schindler: "Every other consideration seemed to be of secondary importance." In December 1826, the composer wrote to his publishers, B. Schott's Sons, asking them not to bring out the Missa Solemnis until he had sent them the metronome markings for it. "In our century, such indications are certainly necessary. Moreover, I have received letters from Berlin telling me that the first performance of the [Ninth] Symphony was received with enthusiastic applause, which I ascribe largely to the metronome markings. We can scarcely have tempi ordinari anymore, since the ideas of unfettered genius must be respected." Beethoven had long been a champion of tempo indications more precise than the conventional Italian designations allegro, allegro ma non troppo, allegro assai, etc. In 1812 or 1813, he met J. N. Maelzel, whose metronome was still in the future, but who had already invented a graduated ticktock device for setting tempo. Beethoven endorsed it eagerly, as we learn from letters and from an article in the Wiener Vaterländische Blätter:
Herr Beethoven looks upon this invention as a welcome means with which to
secure the performance of his brilliant compositions everywhere at the tempi
conceived by him which to his regret have so often been misunderstood.
Maelzel began manufacture of his metronomes in 1816, and Beethoven soon acquired one. He urged their "universal use and distribution." Henceforth, he said, he would attach Italian descriptive terms to a composition only to indicate its "spirit", and provide metronome markings to define its tempo, which is "more like the body." He published Maelzel Metronome figures for the first eight symphonies, the first eleven string quartets, and the septet. He published the "Hammerklavier" sonata, in 1819, with M.M. markings. "Metronome highly necessary", he jotted down in 1825 apropos of a passage in Opus 132. The evidence of his continuing trust in the metronome should be stressed, for at least two reasons. First, Schindler, who thought that he alone held the secrets to true Beethoven performance, sought to discredit the markings, claiming that the composer himself had in later life disowned them, and buttressing the assertion with forged entries in the Beethoven conversation books. (The famous canon "Ta ta ta ta ta... lieber Maelzel", on the Allegretto theme of the Eighth Symphony, may well be another Schindler forgery.) Second, surprisingly few of Beethoven's metronome markings are observed by Beethoven performers. When there is consensus that about half of them are impossibly fast, and a few of them impossibly slow, something has gone wrong. Was it Beethoven's metronome, his employment of it, or his perception of tempo? Or is it present-day interpretation?
The extent of the divergence between markings and interpretation was startlingly revealed by Peter Stadlen in an article in Soundings (1982). He undertook a survey of the nine symphonies as performed on record by eight conductors; of the eleven string quartets as performed by six quartets; of the "Hammerklavier" as played by seven pianists; and of the septet as played by four ensembles. And he determined that of the hundred and thirty-six Beethoven M.M. markings involved only seventy are observed in at least one of the performances. Of the remaining sixty-six markings, sixty-two are faster than in any recording and four are slower.
Was Beethoven's metronome at fault? It seems unlikely. When it went wrong, he had it repaired (as letters of 1819 and 1825 attest). Stadlen worked with metronomes of Beethoven's day and found that they are pretty precise; if anything, they go just a shade too fast—which would result in readings too slow, not (as in most cases) too fast. He tried slowing a metronome down by oiling it with green olive oil adulterated with dust and allowed to oxidize—the most unsuitable lubricant, the British Museum Clock Room told him, that Beethoven could have got hold of in Vienna—and it still kept time. In any case, metronomes that beat too slow also beat erratically, in a way that no musician can fail to notice. So he sought other explanations. He found that the instrument's calibrating scale can be misread by parallax: looking up at a metronome, as from a keyboard, one can read too low a mark on its scale; standing above it and looking down, one too high. But this was still not enough in itself to account for the wider divergences between the markings and their usual execution. Stadlen added the factor of "acoustical delusion." Slow music, he says sounds slower when played on the piano than on strings. And while Beethoven was nearly deaf by 1817 "it is quite conceivable that after a long career as a pianist, his fingers had retained sufficient keyboard sensitivity to compensate instinctively, in such passages, by playing them a little faster while he was determining the tempo of a symphony or quartet at the keyboard." (Another possible acoustical delusion, surely—one Stadlen does not consider—is that when "performing" a piece in one's mind one is apt to set a slightly faster tempo than at a live performance with instruments.) At any rate, by combining (with a delicacy that this summary of his careful work does no justice to) the factors of possible small mechanical error, possible parallax, and possible acoustical delusion, Stadlen determined that sixty of Beethoven's "absurdly fast" figures can be adjusted to be brought within "the realm of the possible." Of the four figures that are "too slow"—all four in the Ninth Symphony—he ascribes one to parallax, one to clerical error, and one to a compound of both. The fourth—the prestissimo at half note=132—he considers to be correct, "and the usually adopted faster tempo... certainly wrong."
Another view is possible, although it takes courage to maintain it against the massed, mutually supportive evidence that Stadlen has collected from so many great musicians, Schnabel, Weingartner, Toscanini, and Furtwängler among them: that Beethoven meant exactly what he wrote, and that our performing traditions are wrong. Benjamin Zander, the conductor of the Boston Philharmonic, argues this case in an essay, "Beethoven's Ninth Symphony and the Metronome", which was distributed to the audience for the performance of the Ninth—an "argument in sound"—that he and the orchestra gave in Carnegie Hall last week.
Tempo and timing matter very much in the Ninth. Its opening, as Tovey says, has been "a radiating point for all subsequent experiments for enlarging the time-scale of music." Beethoven's 1826 letter is testimony to the importance he gave to the metronome markings. But Stadlen's tables reveal that of the sixteen markings in the score Beethoven sent to Berlin only seven are accepted by one or more of the eight conductors in his survey—who are Weingartner, in 1935, and, from the quarter century 1952 to 1977, Furtwängler, Toscanini, Erich Kleiber, Walter, Szell, Kempe, and Karajan (twice). Eight markings are unanimously rejected—four as too fast, four as too slow. (One marking, the finale's brief reminiscence of the opening—allegro ma non troppo, quarter note=88—is, for some reason, unsurveyed.)
For many years, various scores of the Ninth were bedevilled by misprints that clouded discussion of the tempi. The most significant were whole note (instead of half note)=116 at the trio of the scherzo, and dotted half note=96 (instead of 66) at the start of the finale. Back in 1922, the composer C.V. Stanford had pointed out the first error (in a letter to the London Times), yet in 1966 Pierre Boulez essayed the resultant double-speed reading, endeavoring to make one bar of the trio equal to a bar of the scherzo. It's much too fast a tempo; but Beethoven's marking—Stadlen's eight conductors agree—is too slow. (The slowest performance, Furtwängler's, gets down only to 138.) Several conductors pay lip service to the composer's equation of a trio bar with a scherzo half bar; none of the eight surveyed practice it, however. Mr. Zander did, with an almost altogether satisfactory and convincing result. (I say "almost" because it is a long movement to hear at a single pulse.) As for the 96 of the finale: Weingartner worked from a score with this reading and declared it (in his handbook on the performance of the symphonies) too fast for the recitative, though not for the fanfares. The 96 is a simple misprint introduced by the Schott engraver (who was on page 96 at the time). Most conductors remain close to Beethoven's 66.
There were various other little puzzles about the figures. About twenty years ago, I tried to clear them up by consulting, in the German State Library, the pages of the conversation book for the day—September 27, 1826—on which Beethoven and his nephew Karl established the speeds for the symphony. (I had learned of the pages' existence from a passing footnote in Alexander Thayer's biography which no Beethoven scholar—despite all the controversy about the markings—seemed to have followed up. In fact, Stadlen was already at work on them and in an important Music and Letters article in 1967 he published the pages in facsimile, along with the letter to Schott—slightly misrepresented in Emily Anderson's edition of the letters—in which Karl set out the M.M. markings he and his uncle had arrived at.) I compared the pages with the Berlin score into which Karl had entered the markings. And the figures that Beethoven had decided on—however "impossible" they may seem to his modern interpreters—were no longer in doubt. Most of the figures are in Karl's hand, and a few (I think) are in Beethoven's. The pair worked in haste: the next day they were leaving Vienna for a stay in Gneixendorf, and they wanted to have the Berlin copy—a luxuriously bound presentation volume for Kaiser Friedrich Wilhelm III, to whom the symphony is dedicated—ready for dispatch. But their work was careful, and they checked. The last entry (in Karl's hand) reads, "You took it faster than 126./132/That's how we did it this morning." And "Wohin?"—"Where must I enter it?"—Karl asks against the quarter note=80 tempo for the "Joy" theme. He entered it in the Berlin copy at the first sounding of the theme, on the woodwinds, at bar 77 (which is marked allegro moderato in some modern scores—Stadlen follows suit—but is allegro assai in the Berlin copy). He did not reenter the indication—but some modern scores do—when at bar 92 the cellos and basses take up the theme (marked allegro assai in the Berlin copy, but with a pencilled alteration to allegro moderato). All the surveyed conductors then slacken the pace when the singers, led by the baritone, take up the "Joy" theme: Weingartner drops from 100 to 72, Furtwängler from 80 to 54. But Mr. Zander delivered all three statements at Beethoven's quarter note=80, and his triple affirmation made excellent sense.
At the alla marcia, Mr. Zander adopted Beethoven's dotted quarter note=84, and it sounded fine. Both Stadlen and Erich Leinsdorf, in his "The Composer's Advocate", suggest that the dotted quarter note is a mistake for a dotted half note, which would give a tempo twice as fast; but Leinsdorf's "simple conclusion that in some early copy the half note got some ink into its center and became thereby a quarter note" in invalidated by unblotched, unequivocal quarter notes in Karl's hand. Karl was not immune to scribal error, it is true: in the Berlin score and the Schott letter he gives the scherzo unit as half note=116, where only a dotted half note is possible; and in the allegro energico of the finale he omits in the Berlin score—but not in the conversation book or the letter—the dot that his half note=84 requires. His penmanship is immaculate to the point of omitting even necessary spots.
In the prestissimo, Mr. Zander was faithful to Beethoven's half note=132, which Weingartner deemed "impossible... in this passage of boundless enthusiasm" and increased to 152. (The slowest here of Stadlen's conductor's, Toscanini, sets 144, though he drops to Beethoven's mark by the fifteenth bar; Karajan and Szell both adopt 168.) In sum, of the eight disputed M.M. markings in the symphony—those unobserved by any of Stadlen's conductors—Mr. Zander, I reckoned, was faithful to six: the "fast" andante of the slow movement and the "Joy" statement; the "slow" trio, alla marcia, allegro energico, and prestissimo. In the adagio molto, he was much faster than anyone else, though not quite as fast as Beethoven's quarter note=60. For the first movement, marked quarter note=88, he adopted the 76 recommended by Weingartner and practiced by Toscanini and Walter.
Stadlen left it "to future investigations to draw comparisons between metronomized movements that resemble one another structurally or poetically." Mr. Zander, in his essay, notes that the only other adagio molto in 4/4 that Beethoven "metronomized" (nine years earlier), the slow movement of the second Razumovsky, bears the same "too fast" figure—quarter note=60—as the adagio molto of the Ninth, and opens with a theme of similar gait and character. Coincidence? Zander argues that the half notes of both themes are indeed "very slow" at Beethoven's tempo—so slow as to be, at 30, right off the metronome's scale. So "body" (M.M.) and "spirit" (the Italian phrase), he says in effect, are not opposed. But in Stadlen's tables only the Budapest Quartet, in 1951, approaches the Razumovsky marking, and most conductors take the symphony movement at half speed. Arguments about tempo remain abstract until they are put to the test of performance. But I'm not ready, yet, to assess Mr. Zander's Ninth Symphony properly. For one thing, since student days I've been under the influence of a Furtwängler Ninth that—the phrase is trite but true—changed my life, and Furtwängler represents in excelsis the inspired, free, slow-movements-slower, fast-movements-faster approach that has given Beethoven's metronome a bad name. (As a "corrective", I turn to the irresistible Toscanini recording, which treats the markings less extremely but hardly less freely.) For another, alerted by Mr. Zander's essay, I went to Carnegie Hall with a modern, electronic metronome in hand, and full attention—surrender—to the music was inhibited by glances at the instrument's winking light. On the conductor's part, too, there must have been an awareness that the performance was also a demonstration, a making of points. Moreover, I was slightly disappointed by the playing of the Boston Philharmonic, about which I've heard good things, and which I have admired (on tape) in Mahler.
Stadlen, starting from the presumption that half Beethoven's metronome markings—despite his championship of the device—belie his intentions, sought and found explanations of why this should be. Mr. Zander, believing that Beethoven intended what he wrote, sought to realize all Beethoven's markings for the Ninth, not just half of them. His performance, I understand, may be recorded. I hope it is. The points were well and clearly made, but one would need to hear more performances at these speeds before one could find them proved; the weight of familiarity, of years of being stirred by great interpretations at the traditional tempi, is not lightly shed. One would like to hear the other symphonies treated in this way; they contain twenty-one metronome markings faster than any adopted by Stadlen's interpreters. The string quartets contain thirty-four. Other factors need study, too. What effect may our modern, louder instruments, richer-toned playing techniques, and larger concert halls have had upon tempi? All Stadlen's surveyed pianists play the "Hammerklavier" on a modern concert grand, not on a hammerklavier. These matters are not minutiae, for if Mr. Zander is right we have been hearing the music of the greatest composer only in misrepresentation.