Stravinsky’s Rite of Spring is 95 years old this year and is still fresh.

And, sometimes, still controversial. I remember that when Pierre Monteux, who had conducted the riotous premiere in Paris in 1913, conducted it again with the Boston Symphony in 1957, 200 people walked out. Ten years later I heard it played by the Oregon Symphony, and one of Portland’s honorable music critics demanded the next day to know why the city’s intelligent audiences should be assaulted with such noisy, insulting trash.

The Boston Symphony Orchestra, under the baton of Music Director James Levine, will be playing Rite of Spring at concerts on December 4 (open rehearsal at 10:30, regular concert at 8:00 pm) and again on December 5, 6, and 9.

I especially remember the first time I listened to a record of The Rite of Spring while following the orchestra score. Confronting that score was one of the most amazing experiences of my life, a remarkable initiation for a kid of 17 whose previous familiarity with orchestra scores had been limited to miniature scores of Beethoven and Schubert symphonies. It was the first folio-size study score I had ever seen, the Kalmus pirated reprint of the 1923 score published by Editions Russe de Musique, Koussevitzky’s own publishing house.

It starts out with a single staff for high solo bassoon (the solo that caused Camille Saint-Saëns to exclaim,”What instrument is that?”), but already, at the end of the Prelude to Part I, there are 31 staves on the page, mostly divided woodwind in an array of different warbling and twitching parts; there are also glissandi in harmonics for divided violas but you never can hear them. The harmony is sustained below in five solo double basses, also in harmonics. There’s a solo violin playing trills and you never hear that either. Indeed, there are many places in this big score with parts that remain unheard in performance or on records – for instance, the flutter-tongued chromatic scales* occupy nine staves in the score, but I have yet to hear them (high woodwinds).

It is endlessly interesting to study The Rite of Spring in the piano-four-hands score side by side with the orchestral score. Robert Craft’s book says that Stravinsky wrote out both scores at the same time, which accounts for some interesting discrepancies. (The alto flute part, for instance, includes twittering figures in groups of ten 32nd notes; in the piano score, these groups have 12 32nd notes. Why?) One can get a good idea of what details Stravinsky thought were most important, and what was less important – those chromatic scales I just mentioned, for instance, are left out of the piano score.

Most of all, though, what one appreciates in scrutinizing the piano score are two most important elements: what is fundamental rhythmically, and what is fundamental harmonically. The climactic part of “Spring Rounds,”* for instance, uses the full strength of the orchestra ff in chordal layers, with crushing dissonance (including the trombone glissandi – “the vomit,” my teacher, Lukas Foss, called these); but in terms of harmony, one sees how directly Stravinsky must have worked this passage out at the piano, with a purely triadic upper layer, a simple ostinato in parallel fifths in the bass (fortified by tamtam and bass drum), and a third layer in the middle of the texture, of parallel inverted-ninth chords. It’s not in the least atonal; indeed, it is all the more piercingly dissonant because it is tonal. There’s a five-flat key signature, and you can say that it grows out of E-flat Dorian harmony. (The whole section is preceded by six bars with a four-flat signature, over the E-flat trills in the flutes. Why this four-flat signature, I can’t imagine. In the sketches, at the repeat of these bars,* the signature has only two flats – another mystery.)

Indeed, the presence or absence of key signatures can tell you a lot about the way The Rite of Spring was conceived harmonically. (“I had only my ear to guide me,” Stravinsky wrote in 1960. “I heard, and I wrote what I heard.”) The first part of the work to be composed, “Spring Fortunetelling: Dances of the Young Girls,” with the famous F-flat major triad in the cellos and basses and the E-flat seventh in the violins and violas, doubled at the accents by eight horns – this has a three-flat key signature*. You can say that the blank signature in the “Rival Tribes”* really represents C major, despite the complex counterpoint and ostinati; I have always thought of this passage as echoing, more than a little, the Dance of the Tumblers in Act III of Rimsky-Korsakov’s Snegurochka. (Does anyone agree?) Even the “weird chord” in string harmonics that represents the “Kiss of the Earth”* is based on a C-G open fifth.

Only one other place in the orchestra score carries a key signature, in Part II*, the six solo violas in unmistakable B major. This key signature is absent from the parts for accompanying cellos and basses, and one supposes that by this point in the piece, after which Stravinsky’s harmony becomes progressively more polychordal and layered, he decided that key signatures were more of an encumbrance than a help. But the piano score differs. Seven bars* carry the five-sharp signature in the piano score, but not in the orchestra score. Similarly, further on* there is a five-flat signature in the piano score – B-flat minor is plainly in the background when you hear it – but there is no signature in the orchestra score. In the sketches we find that what came out at one section* results from the beginning of the “Sacrificial Dance” “transposed to C-sharp major”. It speaks volumes that Stravinsky thus thought of that famous beginning chord in terms of D major; it has E-flat and B-flat in the bass, but the downbeat is a low D and the upper chord is definitely a D major triad*.

The Rite of Spring is heralded in every textbook as a pioneering work in the evolution of musical rhythm. Most writers point to two examples: the “Spring Fortunetelling” chords*, with an accented pattern. (In this eight-bar phrase, with 32 repetitions of the same chord, the accents are on nos. 10, 12, 18, 21, 25, and 30, if you want to try beating it out literally. This passage, by the way, was among the very first ideas that Stravinsky wrote down when he began to compose The Rite of Spring.) The pattern is repeated with the same chords*. But most remarkably, some 90 bars later, of steady 2/4 in which those same chords have been absent for 67 bars, the pattern of accents reappears complete and rhythmically identical, with entirely different harmony and melody*. This is an uncanny use of isorhythm, a structural technique that flourished and died in the fourteenth century, now reborn in the twentieth. (The composer Frederic Rzewski wrote his undergraduate honors thesis at Harvard, class of 1958, on “The Reappearance of Isorhythm in Twentieth-century Music,” citing noteworthy examples from Berg and Webern.)

The other often-cited example is the beginning of the “Sacrificial Dance,” long considered the most difficult part of the entire Rite of Spring to play and to conduct. (I have played the complete work in public a number of times in two-piano performances and I can tell you that, though it is a good workout, it isn’t that technically difficult; I once played it on the same program with Les Noces, in which I played third piano; and I found Les Noces considerably harder.) The way the beginning of the “Sacrificial Dance” is barred suggests particular conducting patterns to me. The 16th-note is the basic counting pulse – most measures are indicated with 16-note time signatures, 2/8 being 4/16 – but one conducts in eighths (at 126 to the eighth). A downbeat, usually with a low-register D, is indicated gesturally with an accent, a quick bouncing beat, so that the upper-register 16th that immediately follows the downbeat coincides with the upward bounce. This is true as well of bars like the fourth bar (the third 3/16 bar), in which there is no note on the downbeat but only a rest. Beat the 2/8 bar like a normal 2/8 bar, without an accented downbeat; 5/16 bars, like a 2/8 but prolonging the upbeat, like 2+3. What this means in graphic terms is that nearly every bar has an upper-register 16th rest on the downbeat followed by a chord on the second 16th, whatever there may be in the lower register on the downbeat. All of this will suffice until the contrasting section*, when it suffices to beat regular eighths. I still haven’t been able to determine why Stravinsky chose the barring and beaming the way he did in this section. The barring in one section* suggests one pattern of accents; the beaming, an entirely different pattern; and the chords themselves still another. The bars are 3/8, 2/8, 2/8, 3/8, 2/8, 2/8, 2/8 and 3/8; I hear it as though it were four bars of 4/8 plus one of 3/8, with no need to stretch beams over an eighth rest.

In his memoirs (from about 1960) Stravinsky wrote that in 1912 “Diaghilev encouraged me to use a huge orchestra for Le Sacre, promising that the size of our Ballet orchestra would be greatly increased in the following season. I am not sure my orchestra would have been so large otherwise.” In the same memoirs he wrote that the orchestra for the Firebird ballet of 1910 was “wastefully large,” and in preparing the 1919 Firebird Suite he reduced the orchestra, cutting the number of woodwinds approximately in half and removing other instruments. But for The Rite of Spring he called for the largest orchestra he ever used. Woodwinds are in fives with paired auxiliaries (e.g., 3 bassoons and 2 contrabassoons), there are eight horns, and many unusual instruments are called for, including the alto flute, piccolo trumpet, bass trumpet (I wonder why Stravinsky’s didn’t also ask for the alto trumpet, an invention of his teacher Rimsky-Korsakov that he had already used in Fireworks), two Wagner tubas, and some exotic percussion. (Ravel had taken Stravinsky to visit a little shop in Paris that sold Latin-American imports, and Stravinsky had found there a guïro, a notched gourd scratched with a rasp, that found its way into the “Procession of the Sage.”) This is a very large orchestra, certainly, and any group that wants to perform it will have to hire extra players; but it’s not as big an ensemble as Mahler used in several symphonies, or as Schoenberg did in Gurrelieder or Strauss in Elektra. (I think Berlioz’s Grande messe des morts, as he called his Requiem setting, calls for the largest orchestra ever demanded in any work that can be said to be in standard repertory – including 40 extra brass.)

Stravinsky never compiled a separate concert suite from The Rite of Spring as he had from Firebird and Petrushka; the complete ballet is always used in concert performances. Nevertheless, there was a persistent rumor that Stravinsky had wanted to re-orchestrate The Rite of Spring for smaller forces, for practical reasons. Asked about this, Stravinsky denied it; he said that of course he was always touching up parts of the orchestration here and there, but if he ever were to revise the whole thing, he would have enlarged the orchestra by adding harps and saxophones. Imagine what that would have sounded like!

Mark DeVoto, musicologist and composer, is an expert in Alban Berg, also Ravel and Debussy. A graduate of Harvard College (1961) and Princeton (PhD, 1967), he has published extensively on these composers and many music subjects, most notably, music harmony.

Ed: This article posed particular challenges, in that we suspect most people are not reading it with two scores at hand. Therefore, for ease and less frustration in reading, citations of particular section numbers and bar references have been omitted from the body of the text (noted by *s) and appear here, by paragraphs:

Para. 5: between nos. 33 and 37; Para. 7: no. 53, no. 56; Para. 8: nos. 13 through 31, nos. 57 to 66, before no. 72; Para. 9: no. 91, beginning three after no. 101, nos. 111 to 117, no. 167, (see) no. 142; Para 10: no. 13, no. 18, the eight bars beginning at no. 30; Para. 11: beginning at 149, eight bars from no. 159 to 161.