We use cookies to enhance your experience on our website. By continuing to use our website, you are agreeing to our use of cookies. You can change your cookie settings at any time. Find out more


Music in the Early Twentieth Century

CHAPTER 2 Getting Rid of Glue

Satie, Debussy, Fauré, Ravel, Lili Boulanger

CHAPTER 2 Getting Rid of Glue
Richard Taruskin

Richard Taruskin


Who amongst us has not, in his ambitious days, dreamed of the miracle of a poetic prose—musical without meter and without rhyme, flexible enough and sufficiently accented to correspond to the lyrical impulses of the spirit and to the undulations of the world of dreams?1

Charles Baudelaire (1862)

As to the kind of music I want to make, I would like it to be flexible enough and sufficiently accented to correspond to the lyrical impulses of the spirit and to the capriciousness of dreams.2

Claude Debussy (1886)

Ars gallica, the “truly French” art promised in 1871 by the Société Nationale de Musique in reaction to military humiliation at the hands of the Prussians, finally began rather bashfully to show its face toward the end of the next decade. Rather than attempt to vie with the Germans in loftiness and profundity, which merely encouraged “epigonism,” the newer French impulses were at first modest and unthreatening. They aimed at the deflation of rhetoric—an especially pointed gesture in the face of German expressive maximalism—and placed a renewed premium on immediate physical sensation.

The French composers of Strauss’s and Mahler’s generation no longer sought (or no longer said they sought) musically to embody “the Will,” as Arthur Schopenhauer would have said. Rather than use their music to represent and stimulate strong desire—for sexual union, for union with God, or for that mixture of the two known as the “erotic sublime” (less reverently, as “sacroporn”)—they sought to restore “decorative” values to a place of honor. They revived “applied” or utilitarian genres; or rather, they sought to recultivate and reanimate esthetic genres that were based on the utilitarian genres of the past. Rather than a source of power, they sought in music a source of pleasure; rather than the sublime, they sought beauty.

In short, stimulated in part by antagonism toward Germany, in part by an interest in neglected indigenous traditions, and in part by concurrent literary and painterly movements, French musicians began to cultivate a very different sort of modernism from the Germans, and a very different musical technique for embodying it. Like all modernisms, the French version was characterized by a suddenly accelerated rate of stylistic change and innovation. Like all modernisms, it was highly self-conscious, reflective, ironic, and urbane. But where the Germans sought a maximalized emotional or psychological content—which implies a maximum human and expressive “presence”—and would go on seeking it far beyond the point to which we have traced their quest so far, the French modernism that began stirring in the 1880s sought to minimize, and ideally to eliminate, everything the Germans were trying to maximize. To use a term coined around 1925 by the Spanish philosopher José Ortega y Gasset (1883–1955) to describe it in retrospect, the French modernists whom we are about to survey sought “the dehumanization of art.”3

The term can sound rather frightening at the other end of a century so full of social alienation and inhuman deeds. But as Ortega intended it, it had nothing to do with robots or concentration camps. Rather, it stood for an effort to purge art of all those “human, all too human”4 concerns that threaten to turn it into a sweaty, warty human document of only ephemeral value (since emotions are fleeting and desire can be satiated) instead of an elegant or exquisite object of pleasure. “Frivolous!” comes the German retort; to which the French, unperturbed, come right back: “Pretentious!”

To view the opposition as a battle of frivolity vs. pretension is of course to trivialize both positions, as ideological antagonists usually manage to do. Another way of trivializing it, at once more objectionable and more illuminating because it is more alert to the underlying social issues, would be Ortega’s own way. He called the esthetic controversies he was diagnosing in retrospect a war between “two different varieties of the human species,” which he further characterized as “a privileged aristocracy of finer senses” and “the masses.”5 Even more contentiously, he cast the split as one between an art that denied its status as art in the name of irrelevant social ambitions or obligations and an art purified of such impertinences: “an art which can be comprehended only by people possessed of the peculiar gift of artistic sensibility—an art for artists and not for the masses, for ‘quality’ and not for hoi polloi.”6 Or as Anatoliy Lyadov, a Russian composer of the fin de siècle, put it, “Everyone is born with a stomach, but with a soul—one in a thousand.”7

That is going way beyond “Brahminism,” which was a staunchly bourgeois snobbery of education, into “estheticism,” a snobbery of sensibility, ultimately of breeding. The strong manner in which Ortega expressed it reflected the right-wing political attitudes of the 1920s, with their violent recoil against democratic politics. To that extent Ortega’s diagnosis can be criticized as anachronistic. But in milder, less overtly politicized terms the accuracy of Ortega’s diagnosis can be confirmed at the very earliest stages of French modernism: in particular, in a trivial musical genre that did indeed fight pretension with frivolity, and from which French modernism can fairly be said to have taken its bearings.

That genre was the Wagner satire, practiced as a conscious resistance to the dread mage of Bayreuth, even as a sort of exorcism, by a generation of French composers who were, many of them, helplessly in thrall to him in their serious work. In some cases one can juxtapose passages of serious (helpless) Wagner-imitation and frivolous (conscious) Wagner-mockery in the work of a single composer. Take Alexis-Emmanuel Chabrier (1841–94), a founding member of the Société Nationale (hence a proponent of ARS GALLICA) who nevertheless made regular pilgrimages to Bayreuth, worshipped Tristan und Isolde and Parsifal, and could never get out from under Wagner’s thumb as an operatic composer. As a result, his operas have largely vanished from the stage, leaving Chabrier to be represented in active repertoire by a few piano pieces and a single orchestral work, a brilliantly scored exotic “rhapsody” called España (1883).

Example 2-1a, from the Entr’acte before the third act of his comic opera Le roi malgré lui (King in spite of Himself) shows one of the many unconscious—or even worse, perhaps, obsessive—plagiarisms from the Tristan prelude that haunt Chabrier’s work as they do the work of so many of his contemporaries, like César Franck (who, perhaps for this reason, stuck a “poison” label on his Tristan score), Vincent d’Indy (whose opera Fervaal became known as “the French Parsifal”), or Ernest Chausson. Example 2-1b/c, by contrast, show a couple of hilarious spots from Chabrier’s Souvenirs de Munich, a quadrille or suite of ballroom dances for piano four-hands on themes from Tristan und Isolde, composed in 1885–86 after one of his Wagnerian pilgrimages.

Chapter 2 Getting Rid of Glue

ex. 2-1a Emmanuel Chabrier, Le roi malgré lui, III (Entr’acte)

Chapter 2 Getting Rid of GlueChapter 2 Getting Rid of Glue

ex. 2-1b Emmanuel Chabrier, Souvenirs de Munich, refrain (Ecstasies)

Chapter 2 Getting Rid of Glue

ex. 2-1c Emmanuel Chabrier, Souvenirs de Munich, “Death Song”

Wagner spoofs in a similar vein were composed by Gabriel (Urbain) Fauré (1845–1924), André Messager (1853–1929), and others. They always took the form of utilitarian dance pieces, in maximum contradiction to Wagner’s completely “emancipated” and “autonomous” art, and always forced Wagner’s “endless melody” into incongruous clunky cadences after the standard eight or sixteen bars required by the dance figure. The sheer vengeful satisfaction of doing violence to a composer whose music takes such violent possession of one’s responses no doubt contributed to the pleasure of the joke; but perhaps more significant was the reminder that even Wagner’s music is, after all, just music. Putting Wagner’s powerful symbols of desire into contexts where they remain recognizable but are prevented from achieving their uncanny effects reassuringly denatured them, put them in their place. And by extension, art is put in its place—as an enhancer of the quality of life, neither a vicarious lived experience nor a substitute for religion.

That may seem a heavy load of philosophical freight to read into such trifling pieces, and nothing so completely ruins a joke as an explanation, but the same homely genres and the same implicit philosophy (or antiphilosophy) surface in French music around the same time in contexts less directly attached to Wagner. Many French composers, not by conscious collusion but by a shared sense of mission, became preoccupied with similar technical concerns, amounting to a common technical project. That project can be described as one of neutralizing (or perhaps just “neutering”) the Wagnerian desire-symbolism then entering its decadent phase in Germany. The overt Wagner-spoofing was only the jesting public face of a more serious job of exorcism.

To see the project in a more pristine guise we can turn to another set of dances for piano, composed the year after Chabrier’s Souvenirs. Erik Satie (1866–1925) was a twenty-one-year-old Paris Conservatory dropout when he wrote his Trois sarabandes, of which the first is sampled in Ex. 2-2. He was pursuing la vie de Bohème, the “Bohemian life,” in Montmartre (Martyrs’ Hill), the highest point in Paris, then a semirural district where (on account of the steep slopes that had to be climbed on foot) rents were cheap and struggling artists could afford to live. The district’s main industry was its nightlife, and Satie earned his living as the second-string pianist at Le Chat Noir (the Black Cat), a local pub.

Chapter 2 Getting Rid of Glue

fig. 2-1 Erik Satie ca. 1892, by Suzanne Valadon ( 1867–1938).

A complete nobody as far as the musical establishment was concerned, dubbed “the laziest student in the Conservatoire”8 by his exasperated piano teacher, Satie maintained some small notoriety by applying repeatedly (upon learning of the deaths of distinguished members like Charles Gounod or Ambroise Thomas, the Conservatory’s director) for membership in the Académie des Beaux Arts, France’s most prestigious artistic honor society, just as a way of riling the surviving members. He was, in short, a “countercultural” type. Eighty years later he might have been called a hippie.

Chapter 2 Getting Rid of Glue

ex. 2-2a Erik Satie, first Sarabande for Piano, mm. 1–21

Chapter 2 Getting Rid of Glue

ex. 2-2b Erik Satie, first Sarabande for Piano, mm. 100–end

Satie was not the first French post-Wagnerian to write a sarabande. Proclaiming one’s disaffection for the “music of the future” by making an end run around the recent past was already a time-honored—even a shopworn—strategy among the would-be Wagner resisters of the Société Nationale. The Society’s very first concert in 1871 had included a suite for piano by Alexis de Castillon (1838–73), the organization’s first secretary, called Cinq pièces dans le style ancien (“Five pieces in the olden style”). By the end of the century just about every French composer had a work “dans le style ancien” in his or her portfolio: Le roi s’amuse, six airs de danse dans le style ancien for orchestra by Delibes (1882); Suite dans le style ancien, in D, for wind septet by d’Indy (1886); Pièce dans le style ancien for piano (1893) by Cécile Chaminade (1857–1944), to name just a few. The most famous composition of this type was Saint-Saëns’s Septet for trumpet, string quintet, and piano (1881). It did not carry the explicit “olden style” label, but its contents were ostentatiously archaic: “Prélude,” “Menuet,” and so on, and, to conclude, a “Gavotte en final” that contained a bit of fugato, and sported a theme that parodied the sort of leaping, “string-crossing” melodies one found in old violin music (Ex. 2-3).

Chapter 2 Getting Rid of Glue

ex. 2-3 Camille Saint-Saëns, Septet for trumpet, string quintet, and piano, “Gavotte en final”

Satie’s Trois sarabandes differed from these efforts, however. While faithfully cast in old “baroque” dance forms (two or three repeated “strains,” with repetitions fully if needlessly written out rather than marked with repeat signs), their musical style was at once more up-to-date and more pseudoarchaic, showing how notions of the ancient and the ultranovel had been joined in an anti-Wagnerian amalgam. The latest novelty in French music was the “consonant” seventh or ninth chord, in which tones normally treated as dissonances in need of resolution functioned instead as sensuous enrichments of an ordinary triad. Chabrier’s music, especially, luxuriated in this effect. A passage from the prelude to Le roi malgré lui (Ex. 2-4), first performed in May 1887, echoes clearly in the opening phrases of Satie’s first Sarabande, composed four months later.

Chapter 2 Getting Rid of Glue

ex. 2-4 Emmanuel Chabrier, Le roi malgré lui, Prelude

The second chord in the Chabrier passage contains a ninth that is both approached and quitted by leaps, and a seventh that is approached and quitted by chromatic inflection rather than actual resolution to a different scale degree. In the fourth measure, a dominant-ninth chord on C♯ moves in strict parallel motion to another chord of identical structure, showing the sevenths and ninths to be (functionally speaking) every bit as consonant as the thirds and fifths; and the insouciance with which Chabrier continues the parallel fifths and sevenths through another progression shows to what extent he conceived of his harmonies as sheer “sonorities,” altogether apart from any vestige of linear voice leading.

The chords themselves may be regarded as “Wagnerian”; the love music from Tristan und Isolde, to pick only the most obvious example, is saturated with them. But the whole point of Wagnerian harmony was the prolongation of dissonance to the point of pain—a pain arising precisely out of the thwarted need for resolution. Moving such chords in parallel à la Chabrier, or leaping from ninth to ninth, denies and eventually neutralizes this need. As formerly dissonant chords become consonant through such uses (or abuses, as any conservatory professor would contend), the “cadential imperative” is weakened—and with it, the power of music to represent desire.

The closest antecedent to Chabrier’s passage was not in Wagner but in the Coronation bells from Modest Musorgsky’s opera Boris Godunov, where the oscillation of two dominant-seventh chords with roots a tritone apart effectively cancelled the need for that tritone to resolve. It is altogether likely that Chabrier’s harmonic fancy had been stimulated by Musorgsky’s experiments. Saint-Saëns had brought back a vocal score of Boris Godunov from a Russian tour in 1874, and thinking it no more than a curiosity, showed it around to his friends, some of whom it unexpectedly captivated with its revelation that there could still be music that was neither Wagnerian nor anti-Wagnerian, but simply a-Wagnerian. Satie stands in this line of reception. Every chord in the first seven measures of Ex. 2-2a contains at least a seventh, and five of them contain a ninth as well. But not one of these intervals resolves according to traditional rules of voice leading. They are harmonically stable, making the music they inhabit harmonically static.

The B♭♭ major chord in m. 8 is the first unsullied triad in the Sarabande. That must be what gives it its cadential quality. Surely it is not the progression that leads to it that marks it so. The bass moves by fifth, all right, but the leading tone—it would have been A♭ rather than A♭♭—has been suppressed. At a time when German composers were making a fetish out of half-step relations, and when (in the sarcastic words of a Russian critic) a character in an opera “cannot ask for a glass of water without using a fistful of sharps or flats,”9 this French composer was purging his music of functional semitones. In effect, he was ridding his music of its harmonic glue.

Of course, sharps and flats remained in use: depending on the key, they are needed for diatonic as well as chromatic music. And as sheer notational features—rather than emblems of emotional intensity, as the Russian critic implied—they proliferated outlandishly in these early works of Satie. Satie’s penchant for overly complicated note-spellings is probably best viewed as an aspect of the preference for the esoteric and the recherché that we have already learned to associate with “decadence.”

But the truly subversive aspect of the Sarabande was not the superficial outlandishness of its appearance. The truly ticklish thing about it was that all of its important cadential functions have been tonally denatured. The dominant triad in m. 20, the normal “binary” half-cadence, is preceded by a B♭-minor triad: merely a “ii” chord rather than a “V of V.” The final cadence in the piece (Ex. 2-2b) very demonstratively replaces the leading tone (G in m. 100) with a “flat seventh” (G♭ in m. 101, held over in the next bar as the third of a minor-seventh chord on the fifth degree) that maintains the ban on leading-tone resolutions (and hence the suppression of a true dominant function) to the bitter end.

This kind of harmony was often described as “modal” and compared with that of the French music of the sixteenth century and earlier—a music that just then, and not at all by coincidence, was starting to be published in quantity, especially by Satie’s contemporary the nationalistic antiquarian Henry Expert (1863–1952) in a huge series called Les maîtres musiciens de la Renaissance (“The master musicians of the Renaissance”). In fact, however, it was only a pseudomodal style and it was altogether modern. (It, too, had a Russian counterpart in the folk song harmonizations of Balakirev and other Russian nationalists; what Balakirev sought among the peasantry the French were seeking in their musical past—namely, novelty that could claim the pedigreed authority of “authenticity.”) In actual practice, even when the “medieval” or “church” mode in which a piece of old polyphonic music was written contained no leading tone, the leading tone was nevertheless supplied at cadences, just as it is in the modern “harmonic minor,” by applying the rules of what was known as musica ficta. The chaste, charmingly antiquated cadences of Balakirev and Satie conformed to no ancient model. They were a classic case indeed (or, as it came to be called, a “neoclassic” case) of the new passing itself off as old. And for a final irony, this newly manufactured “modal” harmony was quickly and widely adopted as the “authentic” French manner of harmonizing the newly revived and popularized Gregorian chant.

The stable sevenths and ninths and the “modal” cadences in Satie’s Sarabande were two aspects of a single effect, that of purging the music of desire. The leading-tone progressions that filled German music with emotional strain, and that were proliferating like kudzu in the music of Mahler and Strauss, were inhibited in the new French music just as the harmonic texture was being enriched. Where the maximalism of Mahler and Strauss gave one a case—as Strauss’s father complained, according to his son’s famous boast—of bugs in one’s pants, the sonorously opulent yet harmonically inert atmosphere of the Sarabandes “imbue the music,” in the well-chosen words of Alan M. Gillmor, Satie’s biographer, “with a timeless calm.”10

Satie’s next step was to purge the music of that rich harmonic texture and rely on the suppression of leading tones to make possible a “new diatonicism”—music that despite (or even because of) the virtual absence of sharps and flats seemed not merely artless but strangely fresh and rare, as if stripped of memory. What became Satie’s most famous composition, the Trois gymnopédies for piano (1888), was of this type. The curious name was another pseudoclassical affectation. Satie probably found it in a popular music dictionary of the time, such as Dominique Mondo’s Dictionnaire de musique (Paris, 1839), which defined gymnopédie (from the Greek gymnopaidia) as “a nude dance, accompanied by song, which youthful Spartan maidens danced on specific occasions.” (The definition has been traced back to Jean-Jacques Rousseau’s Dictionnaire de musique of 1768.) The slow waltzes that Satie came up with in response to this description surely bore scant resemblance to any ancient model, but they bequeathed a minor genre to later composers (Peggy Glanville-Hicks and John Adams, to name two) who have occasionally written “gymnopedies” on the Satie model.

All three Gymnopédies begin with “vamps”—accompaniments awaiting their tunes—consisting of a pair of chords in a simple alternation suggesting tonic and dominant (Ex. 2-5). In all three cases, however, the tonal “functionality” of the progression is attenuated. In the first, the two harmonies are both “major-seventh” chords, so that a constant level of mild “stable dissonance” is maintained. In the second, the ostensible “I” chord has a sixth (E) in place of its fifth (D), and the ostensible “V” is a minor seventh, stripped of its leading tone (hence of its potency as a dominant). In the third, if the first chord is taken as the tonic, the second can only be construed as some sort of weird “minor (hence not a true dominant), devoid of a leading tone and with its seventh in the bass.

Chapter 2 Getting Rid of Glue

ex. 2-5a Erik Satie, Trois gymnopédies, no. 1, mm. 1–8

Chapter 2 Getting Rid of Glue

ex. 2-5b Erik Satie, Trois gymnopédies, no. 2, mm. 1–8

Chapter 2 Getting Rid of Glue

ex. 2-5c Erik Satie, Trois gymnopédies, no. 3, mm. 1–8

By the time the tune enters in the third Gymnopédie, the second vamping chord is abandoned, to return only at other vamping spots (one in the middle, the other at the end). It is thus exposed as a completely arbitrary sonority without any “inherent” or mandated tonal function. Just as arbitrary is the occasional light chromaticism that impinges from time to time on the strictly diatonic melody. And so, the composer seems to suggest, so are all of our familiar chords, even the ones we consider basic to our “tonal system.” We respond to their functional relationships only because we are conditioned so to do. But we could free ourselves from that conditioning if we wished, and view the field of harmony afresh. Gentle and pretty—and innocuous—though they seemed, Satie’s little dances for piano were radical stuff. They already fully exemplified the esthetic position that Ortega, the later theorist of the avant-garde, would spell out in the 1920s in his famous “seven points.”11 Rather than attempting to provide its audience with a vicarious emotional or spiritual life, the ascetic “new artistic sensibility” Ortega described tended:

  1. 1. to dehumanize art

  2. 2. to avoid living forms

  3. 3. to see to it that the work of art is nothing but a work of art

  4. 4. to consider art as play and nothing else

  5. 5. to be essentially ironical

  6. 6. to beware of sham and hence to aspire to scrupulous realization

  7. 7. to regard art as a thing of no transcending consequence.

When one considers that these early benchmarks of anti-Teutonic modernism were written ten years before the death of Brahms, Satie’s little Sarabandes and Gymnopédies can seem, despite their primitive innocence bordering on infantilism, positively amazing.


(1) Charles Baudelaire to Arsène Houssaye; Baudelaire, Oeuvres completes (Paris: Pléiade, 1956), p. 291.

(2) Claude Debussy to Eugène Vasnier, 19 October 1885; quoted in Stefan Jarocinski, Debussy: Impressionism and Symbolism, trans. Rollo Myers (London: Eulenburg Books, 1976), p. 172n50.

(3) Cf. José Ortega y Gasset, La Deshumanización del arte e Ideas sobre la novela (Madrid: Revista del Occidente, 1925).

(4) Ortega y Gasset, The Dehumanization of Art, trans. Helene Weyl (Princeton: Princeton University Press, 1968), p. 12.

(5) Ibid., p. 6.

(6) Ibid., p. 12.

(7) Quoted in Andrey Nikolayevich Rimsky-Korsakov, “Lichnost’ Lyadova,” Muzïkal’nïy sovremennik, Vol. II, no. 1 (September 1916), p. 33.

(8) Alan M. Gillmor, Erik Satie (New York: Norton, 1992), p. 10.

(9) Vladimir F. Odoyevsky, Literaturno-muzïkal’ noye naslediye (Moscow: Muzgiz, 1956), p. 343.

(10) Gillmor, Erik Satie, p. 37.

(11) Ortega y Gasset, The Dehumanization of Art, p. 14.

Citation (MLA):
Richard Taruskin. "Chapter 2 Getting Rid of Glue." The Oxford History of Western Music. Oxford University Press. New York, USA. n.d. Web. 18 May. 2024. <https://www.oxfordwesternmusic.com/view/Volume4/actrade-9780195384840-chapter-002.xml>.
Citation (APA):
Taruskin, R. (n.d.). Chapter 2 Getting Rid of Glue. In Oxford University Press, Music in the Early Twentieth Century. New York, USA. Retrieved 18 May. 2024, from https://www.oxfordwesternmusic.com/view/Volume4/actrade-9780195384840-chapter-002.xml
Citation (Chicago):
Richard Taruskin. "Chapter 2 Getting Rid of Glue." In Music in the Early Twentieth Century, Oxford University Press. (New York, USA, n.d.). Retrieved 18 May. 2024, from https://www.oxfordwesternmusic.com/view/Volume4/actrade-9780195384840-chapter-002.xml