Chapter Three—
The Songs

By far the largest proportion of Schoenberg's early works consists of Lieder. There survive autographs for thirty-two complete songs (and more fragments) that were written before the end of 1900 but did not appear among Schoenberg's first published collections, opp. 1,2, and 3 in 1903–4.^[1] Since fewer than a third of the manuscripts were actually dated by Schoenberg (only after 1897 did he begin to provide dates with some consistency), the problem of establishing a firm chronology is even harder than in the instrumental works. First attempts along these lines were made by Walter Bailey (1979) and Ulrich Thieme (1979, 127–43), who (independently of each other) separated the early songs, principally on the basis of style, into three main periods: a largely autodidactic phase up to about 1894; a distinctly Brahmsian phase up to about 1897; and a more chromatic, Wagnerian phase from 1897 to 1899. The development thus charted is plausible and agrees essentially with that outlined by Schoenberg himself in his writings.

A more precise chronology, based principally on paper types and handwriting in the song manuscripts, has been proposed more recently (1989) by Christian M. Schmidt in the critical report for SW B1/2/I:42–52. The results are persuasive and are basically consistent with the stylistic development suggested by Bailey and Thieme, as well as with the picture of the early Schoenberg that has been drawn thus far in part I of the present study. In the absence of precise datings from Schoenberg himself or from other biographical sources, Schmidt's work should be accepted as the most accurate available; it forms the basis of the list of early songs through 1897 given in table 2.

[1] The early songs, which are scattered among several collections and institutions, were first inventoried by Stein 1977 and Bailey 1979, whose work has now been superseded by, and incorporated into, the comprehensive critical report prepared by Christian M. Schmidt for SW B½. Seven of the early songs were published as Schoenberg 1987. All have now appeared in SW A1 and A2.

---

---

Based on a fresh assessment of these works, and in light of the more precise chronology now available, I would refine and reshape somewhat the version of Schoenberg's development given by Bailey and Thieme. We can discern a fourstage development, which forms the basis for the analytical commentary in this chapter and can be summarized as follows:

1.  Most of the songs that appear to antedate 1894 are rather pallid imitations of Brahms and Schumann, with one exception, Schilflied ("Drüben geht die Sonne scheiden") of 1893, which shows a highly original, if still tentative, compositional voice.

2.  With Ecloge of 1895, which went through three different versions (more than any other surviving early song), Schoenberg breaks free of the the more obvious Brahmsian imitation and refines certain harmonic-formal techniques explored in Schilflied.

3. With the two Heyse settings Mädchenlied and Waldesnacht, which were almost certainly composed in 1897, Schoenberg seems to return more self-consciously to Brahms, but now achieves a genuine internalization or absorption fully on a level with that of the contemporaneous D-Major Quartet.

4. Schoenberg moves definitively away from the Brahms idiom—but not from its fundamental compositional principles—in the two Dehmel settings from the fall of 1897, Mädchenfrühling and Nicht doch! Here he reaches toward the newer, more individual kind of musical expression that was to culminate in his works of 1899.

One aspect of Schoenberg's activity as a song composer that appears quite early on (although it is by no means unique to him) is a tendency toward concentrated involvement with specific poets for short periods. Moreover, the quality of the poetry seems directly related to the musical results. Thus, in the earliest period, up to about 1895, there are eleven relatively uninspired complete (and other fragmentary) settings of poems by Ludwig Pfau, a minor figure from the sentimental, volkstümlich romantic tradition. With the two Heyse settings and then the Dehmel settings of 1897, the higher quality of the poetry seems to stimulate a higher level of musical composition.

Schilflied

Schilflied, which Schoenberg's childhood friend David Josef Bach has testified was composed in 1893 (Bach 1924, 317), brought Schoenberg his first semi-

---

official recognition: it was awarded a composition prize by the amateur Viennese orchestra Polyhymnia, the group Schoenberg was to join as cellist in 1895.^[2] In another song datable to 1893, In hellen Träumen, Schoenberg worked with a rather mawkish text by his schoolmate Alfred Gold. In Schilflied, he turned to a more venerable poem, one of a set of five Schilflieder published in 1832 by the romantic author Nikolaus Lenau. (Although Schilflied thus refers as a title to the set of poems, and not the individual text, which begins "Drüben geht die Sonne scheiden," I shall for ease of reference continue to call Schoenberg's song Schilflied in the present chapter.) This group of poems became extremely popular with composers of the nineteenth century. A scan of Ernst Challier's Grosser Lieder Katalog reveals that by the time Schoenberg wrote his song, in 1893, at least fifty-seven settings of this particular text, "Drüben geht die Sonne scheiden," had appeared in print (Challier 1885). Conceivably the choice of text was dictated by the Polyhymnia prize competition.

The poem (printed here as it appears in Schoenberg's manuscript) is in many respects a characteristic creation of German romanticism, in which nature becomes a metaphor for, or projection of, the feelings of a solitary protagonist:

Drüben geht die Sonne scheiden,
Und der müde Tag entschlief.
Niederhangen hier die Weiden
In den Teich, so still, so tief.

Und ich muß mein Liebstes meiden:
Quill, o Träne, quill hervor!
Traurig säuseln hier die Weiden,
Und im Winde bebt das Rohr.

In mein stilles, tiefes Leiden
Strahlst du, Ferne! hell und mild
Wie durch Binsen hier und Weiden,
Strahlt des Abendsternes Bild.

[2] As has been realized by Schmidt, the entry for Schilflied in Maegaard 1972, 1: 156, is in error in stating that the manuscript is part of group of Lieder (at the Morgan Library) enclosed in a wrapper headed "4 Lieder korrigiert 16." This information led Bailey (1979, 53–54) to suggest that the surviving manuscript of Schilflied may represent a later revision. The wrapper actually belongs to drafts of four songs from the Hanging Gardens cycle, op. 15, also at the Morgan (nos. 3, 4, 5, and 8; see Maegaard, 1: 62, and SW B1/2/I: 24 and 50–51). (Schmidt seems to accept that the wrapper reads "korrigiert" and thereby acknowledges a potential problem with the Hanging Gardens songs, for which a 1916 revision is unlikely. To my eye, the scribble is by no means definitively readable as "korrigiert.") In short, there is no physical evidence to contradict Bach's suggestion that Schilflied is a song of 1893.

---

Over there the sun goes down, and the weary day closed its eyes. Here the
willows droop in the still, deep pond.

And I must stay away from my beloved: flow, my tears, flow forth! Here
the willows are rustling in sadness, and the reed trembles in the wind.

Into my still, deep suffering you shine, O distant one! Bright and gentle, as
through rushes here and willows, shines the image of the evening star.

Schoenberg shapes Lenau's three stanzas into a rounded four-part song, in which the central stanza is divided musically into two parts: the overall form is thus A (m. 1), B (m. 19), C (m. 27), A' (m. 35). The song represents perhaps the earliest surviving example in Schoenberg's works of the kind of motivic-rhythmic concentration that was to become a hallmark of his compositional technique and that we have noted in the D-Major Quartet and F-Major Scherzo.

The prelude (Appendix ex. D) is dominated by an almost obsessive repetition of a two-note, short-long figure, which is traded between the two hands. When the voice enters in m. 5, the motive continues in the accompaniment, but its articulation is now modified. The left hand plays as before, but in the right, slurs now join the longer note to the succeeding shorter note. The motivic forms in the voice at the end of the stanza, in mm. 11–13, are clearly heard as further modifications of the original two-note figure.

The economy of the motivic language may well have been inspired by Schoenberg's early encounters with Brahms's music, but there is in fact little else of a recognizable Brahms style in Schilflied. In the harmonic language, especially, Schoenberg seems engaged in an empirical or intuitive exploration of the kind of chromatic chords he was later to label "vagrant." Extensive use of vagrant harmonies is, of course, a major aspect of the style of Liszt, Wagner, and Wolf, a repertory that Schoenberg claimed to have discovered only in 1898–99 (as reflected in works discussed in succeeding chapters). Schilflied in no way invalidates that claim, for in it the augmented, diminished, and half-diminished chords are employed in a manner very different from that of Wagner or Wolf. These harmonies do not arise primarily from the kind of stepwise chromatic voice-leading that might be called Tristanesque. Instead, they occur to extend, delay, or obscure basic diatonic progressions. Hence my characterization of Schoenberg's harmonic language here as "empirical" and intuitive; it is as if he is discovering the properties of vagrant chords on his own.

The prelude essentially sustains or prolongs a dominant, E-major harmony. However, as I have indicated with the brackets underneath mm. 1–4 in Appendix ex. D, the bass line actually arpeggiates not the dominant triad, but a more ambiguous half-diminished seventh chord,

. The first vocal phrase carries the ambiguity a step further: it ends in m. 8, not with the expected dom-

---

Example 3.1
Tonal structure of Schilflied.

inant triad, but with an augmented chord spelled

. The C can be understood as an upper neighbor to the implied B of the dominant triad; but the C conspicuously fails to resolve and is instead carried over in the harmonies of the first measure of the next phrase. The conclusion of the A section is also colored by the augmented triad, which evolves by chromatic motion into two other vagrant harmonies under the repeated E of the voice in mm. 12–13.

In this first section of the song, then, Schoenberg uses the vagrant harmonies to deemphasize the tonic-dominant axis and thereby capture something of the general Müdigkeit or lassitude conveyed by the poem. In the impassioned B and C sections, the vagrant harmonies surge more forcefully and continuously. The initial key area of the B section, D minor, is immediately overridden by a succession of seventh chords: minor, half-diminished, and diminished. Particularly striking here is the tritone motion in the bass of mm. 22–24, supporting the juxtaposition of a C-minor seventh chord and a second-inversion augmented triad with a

root and bass. Only in m. 25 is a stable G-minor triad reached. The more tranquil C section is oriented around F major; just before the fermata of m. 34, the F appears in the bass to establish itself as the root of a German sixth chord in A minor and usher in the reprise.

If we step above, or behind, the many harmonic regions touched on in Schilflied, what emerges is a coherent, if idiosyncratic, large-scale structure, which is sketched in ex. 3.1. The song proceeds essentially by fifth through the arrival on G minor in m. 25. The status of G minor as the long-range goal of this "progression" helps to explain why it is given such emphasis in mm. 25–26—more emphasis, indeed, than is given to the tonic in the A section. At the end of the C section, this G can be heard to descend stepwise to F, or the sixth degree in A minor, which in turn drops to E to begin the A' section.

Because of the tension sustained in this way across large spans, the song can even be heard as having some of the tonal dynamic characteristic of a sonata form: the A section functions as a kind of exposition, centered around tonic and dominant; the B section is a far-ranging "development," reaching a remote G minor; C is a "retransition," moving to a sonority that becomes recognizable as the sixth degree, upper neighbor to the dominant; A' functions as a "recapitulation." I use these terms not to imply that Schoenberg was attempting a song in

---

sonata form, but to suggest that the song manifests ambitious "instrumental" impulses that might be said to strain against the traditions of the song medium.

Ecloge

Some of these same impulses can be detected in another remarkable song, Ecloge, which, although not dated by the composer, was probably composed in 1895. The poet, not named on any of the three manuscripts for the song, has recently been identified (SW A2: 67) as the Czech Jaroslav Vrchlický (a pseudonym for Emil Frida). Schoenberg appears to have taken this text from a German translation of some of Vrchlický's lyrics that appeared in 1895.^[3] The poem is not an eclogue in the strictly classical sense of a pastoral dialogue, but is rather an attractive romantic lyric associating spring, described in stanzas 1 and 2, with love, described in parallel terms in stanzas 3 and 4:

Duftreich ist die Erde und die Luft krystallen,
Und das Moos erzittert unter deinem Fuß,
Aus dem Schilfrohr hör'ich's wie von Pfeifen schallen,
Und vom Hagedorn fälk heller Blütengruß.
Und das Aug' von Freude naß,
Fragst du: Ja, was soll all das?
"Was?"
Ruft der Vogel und die Blume spricht:
"Anders kommen doch des Lenzes holde Wunder nicht!"

Hell dein Blick, dein Atem süß vom Duft der Erlen,
Und es bebt dein Busen, wie ich dich umfang';
Wie aus hartem Felsen springen Quellenperlen,
Bricht aus meinem Herzen glühender Lieder Drang.
Und das Aug von Freude naß,
Fragst du: Ja, was soll all das?
"Was?"
Ruft der Vogel und die Blume spricht:
"Anders kommen doch der Liebe holde Wunder nicht!"

[3] See Gedichte von Jaroslav Vrchlický, translated by Friedrich Adler (Leipzig: Reclam, n.d.). Since Adler's preface is dated December 1894, we can assume the volume appeared in 1895. This book was only one of several collections of Vrchlický's poems published in German, beginning in 1886. The poem "Duftreich ist die Erde," no. 5 in a volume of ten eclogues originally published in Czech in 1880, appeared in several of the German collections; but only the Adler translation of 1895 matches the one used by Schoenberg (with minor differences of spelling and punctuation). Thus we can safely assume (with Schmidt, SW B1/2/I:11) that this volume served as Schoenberg's source. The date of 1895 also fits in plausibly with the stylistic-compositional development traced in this chapter. The text of the poem as printed here is taken from the Adler volume. In the Adler volume the generic title "Ekloge" is given with a k, whereas in the only one of Schoenberg's manuscripts to bear the title at all (at the Schoenberg Institute), it appears with a c. In SW, Schmidt follows Adler's spelling; here I have followed Schoenberg's, hence Ecloge.

---

The earth is rich with scents, the air crystalline, and the moss trembles
under your foot. From the reed I hear a sound like that of pipes, and from
the hawthorn descends the bright greeting of blossoms. What does all this
mean? you ask, your eyes wet with joy. "What?" calls the bird, and the
flower responds: "The sweet miracle of spring comes in no other way."

Your gaze is bright, your breath sweet with the scent of alder, and your
breast throbs as I embrace you. As pearls of water spring forth from hard
stones, thus does a rush of radiant songs burst from my heart. What does all
this mean? you ask, your eyes wet with joy. "What?" calls the bird, and the
flower responds: "The sweet miracle of love comes in no other way."

Following the poem, Schoenberg's setting is strophic on the largest scale: it is based on an almost exact repetition of a musical unit of 49 measures. The way in which the musical strophe is shaped—and, among the three manuscripts, reshaped—bears certain resemblances to Schilflied. The overall form of the strophe can be considered A (mm. 1–15), A' (16–23), B (24–30), A" (31–49) (see Appendix ex. E).

As in Schilflied, Schoenberg attempts in this song continually to sidestep the tonic,

, which is reached in only m. 4 as the last component of a sequence moving (as shown in the Appendix). Because arrives in this way, with no special emphasis or preparation, we may not even recognize it as the tonic. Nor is it confirmed in what follows. The vocal part in the A section breaks off in m. 12 in midstream, as it were, on a chord. In the two-bar transition between sections, mm. 14–15, Schoenberg again treats the tonic simply as part of an ongoing sequence, ii⁷ -V-I-IV, leading to the return of A'.

The broad, dissonant leaps make the vocal part of Ecloge considerably more advanced than that of Schilflied. The piano part also has a more independent, continuous motivic process, which unfolds in the tenor range, underneath the cascading eighth notes. As in the other song, the basic motive is a very concise one, here the four-note

figure of m. 1. Measure 2 presents a rhythmic variant of this motive and serves to mark off a basic two-bar unit, which is then repeated sequentially. At the respective conclusions of the four-bar groups in mm. 4 and 8, the motive is further transformed into a three-note figure, quarterdotted quarter-eighth. This in turn becomes the sole motive form to appear in mm. 9–12, where Schoenberg builds to the conclusion of the A section.

The virtually equal status of piano and voice in Ecloge is confirmed by the exchange at the start of the A" section. In mm. 31–34, the voice takes over the the-matic skeleton of the right-hand eighth-note figuration of mm. 1–4, and the right hand of the piano plays the disjunct theme formerly sung by the voice. The exchange ends in m. 35 as Schoenberg introduces new vocal material and a new kind of texture for the line beginning with "Anders kommen." It is here that we find

---

the first real signs of technical unsureness in the song. The passage seems clumsy, unmotivated, perhaps because we have not had enough of a real "return" to justify a new departure, which is then followed by a sudden push toward the apparent climax of the song on the big

chord and high of "Wunder" in m. 42.

My own dissatisfaction with the passage beginning at m. 35 was apparently shared by Schoenberg, who altered it in each of the three autographs for Ecloge. Although none of the three manuscripts bears a date, their order of completion can be surmised on the basis of the passage in question. The earliest version is a fragment (in the Nachod collection at North Texas State University; reproduced in Kimmey 1979, 204–6), which breaks off after the first musical strophe (and in which Schoenberg has inadvertently substituted "der Liebe," from stanza 2, for "des Lenzes" from stanza 1). The version of the song at the Pierpont Morgan Library (as reproduced in Appendix ex. E), close in handwriting style to the Texas copy, is almost certainly the second in order: it is a complete draft of the song, but shows signs of intended revision in mm. 24–25 and 35–37, which have wavy lines drawn across them. The copy at the Schoenberg Institute is the third manuscript. A fair copy with breathing marks written in, it was apparently intended for performance.^[4]

Originally, the Schoenberg Institute copy (the third) corresponded almost exactly to the Morgan one (the second), but Schoenberg began to revise extensively at precisely one of the spots marked with a wavy line in the Morgan copy, at "Anders kommen" in mm. 35ff. Although the crossots and pasteovers in the Institute copy do not in fact yield a complete, coherent draft of the song, we can reconstruct at least the state in which he left the "Anders kommen" passage before apparently abandoning work on the song. The three versions of this passage are superimposed in ex. 3.2.

In Schoenberg's first attempt, ex. 3.2a, the parallel-fauxbourdon texture at "Anders" involves even the bass, which is high in its register and plays in eighth notes. The vocal line and the top part of the piano here are initially a fourth higher than in the version of 3.2b. In 3.2b, Schoenberg brings the first "Wunder nicht" (mm. 7–8) down a minor third, to

,  probably so as to avoid usurping the climactic to be reached on the later "Wunder" of m. 12. Text repetition is also handled differently in 3.2a and b. In 3.2a, Schoenberg repeats the line of text intact in mm. 9–13; in 3.2b, he adds still more text, in effect delaying the arrival on the final "nicht."

The most far-reaching changes, however, come in the third version, 3.2c. Two aspects are particularly striking: the different figuration at "Anders" and the new

[4] That this third draft of Ecloge forms part of a gathering with Waldesnacht, which was almost certainly composed in 1897, has led Schmidt to suggest that Schoenberg may have attempted this revision in 1897, thus two years after the first draft (SW B1/2/I: 45).

---

Example 3.2
Comparison of versions of Ecloge.

---

Example 3.2
continued

---

---

harmonization from "Blume" onward. As to the first: Schoenberg must have realized that the parallel chordal style is too anomalous, especially in a recapitulatory section. He thus abandons it and reintroduces the basic slurred eighth-note figuration of the A section. At the same time he reshapes the harmonic progression so as to avoid arriving on the tonic at "spricht" in m. 4. He understood that this tonic made for too much closure in versions a and b and that it tends to undercut the arrival on the tonic

at "Wunder." Moreover, by this point we have heard this particular progression, twice before, at the beginning of the A and A' sections.

Schoenberg thus reharmonizes the

in the voice at "Blume spricht." As shown in ex. 3.2c, he displaces the to m. 4 ("spricht"), preceding it at "Blume" with a neighbor-chord spelled as a seventh (the chord has no functional label in The resolution of the dominant now falls on "Anders," in m. 5, a bar later than in the earlier versions. Rather than closing off harmonic motion, as happened in versions 3.2a and b, the tonic now forms the beginning point of a strong rising bass line that pushes forward more quickly to the secondary dominant on "nicht" and then to the tonic .

In 3.2c, Schoenberg has shifted the weight and placement of the climax. In the earlier versions the

chord had been reached more circuitously, and its power had then been vitiated somewhat too quickly by the immediate melodic cadence to (m. 13). In 3.2b, the extension by word repetition does not change this fundamental situation, since the voice hovers on for the last four measures. But in 3.2c, the voice descends more slowly—and appropriately—to the final which now coincides with the actual arrival of the root in the bass.

I have dwelled on the successive revisions of this small passage from Ecloge because they have a significance broader than their immediate context. They show Schoenberg grappling with fundamental compositional issues of thematic, formal, and harmonic balance. More than in Schilflied, he seems to be questioning the proper shape and status of a "recapitulation" or return: how to prepare it harmonically, how long it needs to be to accomplish its formal purpose, how it should transform earlier material, and so forth.

In these ways, both Schilflied and Ecloge open up a number of promising paths that Schoenberg might have been expected to follow immediately in 1895. In fact, it is not clear just what direction his song composition took over the next one or two years. This was the period of his private study with Zemlinsky, who may have encouraged Schoenberg to return to and master the Brahmsian model, to "get it right," so to speak (something that, as I suggested in chapter 1, Zemlinsky himself failed to do in Heilige Nacht). In any case, this period culminates in Schoenberg's most profoundly Brahmsian songs, Mädchenlied and Waldesnacht (1897). Here we find the kind of assimilation and absorption missing from earlier efforts, especially apparent in the phrase structure, motivic development, and

---

harmonic expansion. Each of the two songs seems to draw consciously on a different Stimmung typical of Brahms—the first light and folk-like, the second slow and broadly lyrical. In both, Schoenberg turns to Paul Heyse, whose poems Brahms himself had set on several occasions, and who is a substantial cut above Ludwig Pfau.

Mädchenlied

Schoenberg sets the poem "Mädchenlied" as a modified strophic song in which two poetic stanzas are grouped into one musical strophe:

Sang ein Bettlerpärlein
Am Schenkentor,
Zwei geliebte Lippen
An meinem Ohr:

Schenkin, süße Schenkin,
Kredenz dem Paar,
Ihrem Dürsten biete
Die Labung dar!

Und ich bot sie willig,
Doch der böse Mann,
Biß mir wund die Lippen,
Und lachte dann:

Ritzt der Gast dem Becher
Ein Zeichen ein,
Heißt's er ist zu eigen
Nur ihm allein.

A pair of beggars sang at the tavern door; two seductive lips at my ear said:

Barmaid, sweet barmaid, serve this pair. For their thirst offer refreshment.

And I offered it gladly, yet the wicked man bit and hurt my lips and
 laughed:

If a guest carves his sign on a cup, they say it belongs to him alone.

The song (first strophe in Appendix ex. F) is a direct counterpart to the first movement of the D-Major Quartet from the same year, not only in key and mood but also in the triadic/pentatonic build of the opening melody, which likewise emphasizes the sixth degree, B, and a B-minor harmony. Brahms, of course, is the ultimate model. Indeed, Schoenberg seems to have had a specific Brahms song in his ear: Ständchen, op. 106, no. 1, whose first phrase concludes with the same

melodic figure, harmonized by vi-I.

---

It is in the more subtle relationships between voice and piano and in the fluid phrase structure that Schoenberg's song reveals a more genuine Brahmsian inheritance. In the first two phrases, mm. 1–5, the accompaniment repeats a brief three-note motive, x, the neighbor-note figure heard first as

. In the second half of the first strophe, mm. 6–7, x migrates from the piano into the voice, at "süße" and "kredenz dem."

This motivic transference forms part of a still more sophisticated exchange of material between voice and piano. As the voice reaches the half-cadence on the dominant in m. 5, the piano begins a restatement of what was originally the vocal theme, now in the dominant. From the second half of m. 6 this restatement begins to deviate slightly from the original (an exact transposition would bring an E in the right hand on beat 3 of m. 6), but the contour and rhythms are fully recognizable through m. 7.

This piano statement actually overlaps with the conclusion of the first vocal phrase in m. 5 and the beginning of the next one in m. 6. There is thus an asynchronous relationship between voice and accompaniment here, perhaps intended by Schoenberg as a kind of musical corollary of the drunken beggars outside the tavern. The voice and accompaniment could be said to get back into phase in m. 8, for the final phrase of the strophe.

The flexible relationship between piano and voice in Mädchenlied makes the analogous techniques in Ecloge —the piano's motivic development and the interchange of material at the recapitulation—look rudimentary. More significantly, the comparison suggests that it was in emulating Brahms's music that Schoenberg found the ways to expand on techniques already present in his own earlier songs.

Waldesnacht

If Mädchenlied pays homage to Brahms's sophisticated Volkston, Schoenberg's Waldesnacht draws impressively on the rich fund of broad, slow songs like Feldeinsamkeit. Where Heyse's "Mädchenlied" poem is concise and epigrammatic, his "Waldesnacht" is expansive, almost hymnic:

Waldesnacht, du wunderkühle,
Die ich tausend Male grüß,
Nach dem lauten Weltgewühle
O wie ist dein Rauschen süß!
Träumerisch die müden Glieder
Berg' ich weich ins Moos,
Und mir ist, als würd' ich wieder
All der irren Qualen los.

---

Fernes Flötenlied, vertöne,
Das ein weites Sehnen rührt
Die Gedanken in die schöne,
Ach, mißgönnte Ferne führt!
Laß die Waldesnacht mich wiegen,
Stillen jede Pein,
Und ein seliges Genügen
Saug' ich mit den Düften ein.

In den heimlich engen Kreisen
Wird dir wohl, du wildes Herz,
Und ein Friede schwebt mit leisen
Flügelschlägen niederwärts.
Singet, holde Vögellieder,
Mich in Schlummer sacht!
Irre Qualen, löst euch wieder;
Wildes Herz, nun gute Nacht!

Forest night, wondrous cool, I greet you a thousandfold; after the noisy
turmoil of the world, oh, how sweet is your rustling! Dreamily I bury my
weary limbs in the soft moss, and it is as if I were freed from all my
confused torments.

Sound, distant flute song, which stirs a vast longing and leads my
thoughts to the lovely distance, oh so begrudged! Let the forest night
lull me and silence my pain, and I breathe a blissful contentedness with its
fragrance.

In your secretive, close confines you will recover, restless heart. And peace
floats downward on gently beating wings. Sing me to gentle slumber,
tender bird songs! Begone, delirious torments; good night, then, restless
heart!

Schoenberg's setting of the poem is modified strophic; the changes come principally in the vocal part of lines 5–6 of each stanza, analogous to mm. 14–17. (Appendix ex. G gives only the first musical strophe in full.) For this song, Schoenberg had a direct model in Brahms's own setting of the Heyse's "Waldesnacht" as a song for mixed chorus, op. 62 (ex. 3.3). A comparison of the opening phrase in the two settings will show how Schoenberg responds to Brahms's predilection for asymmetry. The first two lines of the poem, which have eight regular metrical stresses, would fall naturally into a phrase of four full measures, divided 2+2. Where in one of his earlier songs Schoenberg would almost certainly have followed the four-square approach, here he emulates Brahms in extending the first phrase to five measures, or eighteen quarter beats. But he follows the spirit rather than the letter of Brahms's practice: Brahms augments the note values on "kühle" and "die"; Schoenberg creates a beautiful ascending melisma

---

Example 3.3
Brahms, Waldesnacht , op. 62, no. 3.

on the first syllable of "wunder." By means of word repetition, Brahms extends his second phrase still further, to six full measures, or twenty-four quarter beats. Schoenberg moves in the opposite direction, compressing lines 3–4 into four measures plus an upbeat, or seventeen quarter beats.

Schoenberg's integration of chromatic detail within a diatonic context places this song well beyond his other Brahmsian efforts and can, indeed, stand beside Brahms's own practice in chromatic passages like mm. 6–10 of the choral song Waldesnacht. We might note especially Schoenberg's elegant treatment of the flatted sixth degree,

. This is the first non-diatonic note heard in the song; it appears in the bass of m. 5 to color the word "kühle." The same pitch, enhar-

---

Example 3.4
Schoenberg, Waldesnacht , strophe 3, mm. 58–61 (vocal part only).

monically respelled as

reappears on the corresponding rhyme "wühle", in m. 9, as part of a strong stepwise bass progression. At the cadence in mm. 11–12, the appears in the voice and piano as an appoggiatura to the dominant.

The

becomes the basis for chromatic effects that reach still farther in mm. 14–17, where it forms part of a German sixth chord in the tonic minor, spelled (mm. 14 and 16; the chord has the G in the bass in m. 16). This harmony does not resolve to the dominant; it bypasses in the bass and steps down to a ii⁷ chord at the fermata. The is never actually resolved as such; it is rather "neutralized", as Schoenberg would say, by the return of the diatonic sixth degree in m. 19.

Two other aspects of the song merit comment here. The first is the subtle evolution of the piano texture across the strophe. The staggered arpeggiation of mm. 3–12 gives way in m. 13 to full chords alternating between the hands. After the fermata of m. 17, this texture evolves into the parallel chordal style in which the piano doubles the voice. We have encountered this texture before, in the "Anders kommen" section of Ecloge. There it seemed merely awkward, and Schoenberg eventually abandoned it (see ex. 3.2c above). In Waldesnacht, however, the parallel chordal style has been carefully prepared and makes for a wonderfully effective climax to the musical strophe.

The second aspect of the song that is noteworthy is the transformation of the vocal part in the third strophe, at mm. 58–61 (ex. 3.4; cf. mm. 14–17 of Appendix ex. G). Here the voice in essence takes over the top line of the piano part (with the addition of a downward sixth leap at "Vogel"). The enharmonic and chromatic steps carry this phrase well beyond the Brahmsian idiom of the rest of the song. Indeed, this kind of line—as well as its harmonic underpinning, including a whole-tone chord on the last beat ofm. 14/58—would be at home in a Schoenberg work of 1899, or perhaps in Gurrelieder.^[5]

[5] This vocal line actually represents a revision written by Schoenberg over the original line, which was as in strophe i. In SW A2: 85, Schmidt opts for the original version and banishes the chromatic revision to the critical report (SW B1/2/I: 294). This seems to me an unfortunate editorial decision. Schoenberg 1987 gives mm. 58–61, as in my ex. 3.4.

---

This passage shows quite vividly that Waldesnacht marks the farthest Schoenberg could go within the Brahms "style". Together with the variation movement of the D-Major String Quartet (examined in the previous chapter), the song can be said to represent the culmination of his period of Brahms study; it is his graduation piece, so to speak, or the work pointing the way out of the "Brahms fog". That path was to be opened still more forcefully in the fall of 1897 with Schoenberg's turn away from the kind of poetry represented by Heyse to the newer kind of verse embodied in the works of Richard Dehmel.

The Dehmel Settings of 1897

In one of the early manifestos of German modernism, a polemical essay of 1885 called "Die neye Lyrik", the literary critic Karl Bleibtreu called for a reaction against the folk-like tradition in German poetry as represented by Heyse. He bemoaned the fact that "the Volkslied, coming by a detour through Goethe, is taken by most critical papists as the only valid norm of the so-called 'genuine [echte] lyric.'" He added, with obvious sarcasm, "Ah, the 'genuine lyric' is indeed so easy: a little mood, a little rhyme, and the masterwork is ready." Bleibtreu demanded a new kind of poetry that would exhibit not only greater naturalism (the first catchword among the early modernists) but also greater subjectivity: "In the first place, subjectivity must be released, in order to break the stiffness of the conventional patterns" (in Ruprecht 1962, 51–53).

The new subjectivity or individualism in poetry, strongly influenced by Nietzsche in the late 1880s and 1890s, was to take many forms in expression and technique. But it was nowhere embodied more strikingly than in the verse of Richard Dehmel (1863–1925), whose first collection of poetry appeared in 1891.^[6] By the late 1890s, Dehmel was perhaps the most famous poet in Germany (or with Stefan George, one of the two most famous). The public and press were at once attracted and scandalized by Dehmel's eclectic blend of what the critic Julius Hart in 1896 characterized as "archaism, symbolism and allegory, everyday-realistic naturalism, with elements of sexuality, immorality and Satanism, and the Nietzschean superman" (in Ruprecht and Bänsch 1970, 15).

This poetry apparently held no attraction for the older generation of composers; Brahms dismissed it as "not well suited for musical treatment" (in Birke 1958, 280). But the younger musicians, including Richard Strauss, Zemlinsky,

[6] The principal collections of Dehmel's poems published in the 189os include Erlösungen: Eine Seelenwandlung in Gedichten und Sprüchen (Stuttgart: Göschen, 1891); Aber die Liebe: Ein Ehemanns-und Menschenbuch (Munich: Albert, 1893); Lebensblätter: Gedichte und Anderes (Berlin: Verlag der Genossenschaft PAN, 1895); Weib und Welt: Gedichte (Berlin: Schuster & Loeffler, 1896); and Erlösungen: Gedichte und Sprüche, 2d ed., rev. (Berlin: Schuster & Loeffler, 1898).

---

Reger, Schoenberg, and (slightly later) Anton Webern, were strongly drawn to it.^[7] Dehmel was the first major contemporary poet to whom Schoenberg devoted sustained creative attention. And it was a conversion that was to have farreaching implications for his musical style.

In a well-known letter to Dehmel of 1912, Schoenberg himself acknowledged the poet's profound impact, noting that

your poems have had a decisive influence on my development as a composer. They were what first made me try to find a new tone in the lyrical mood. Or rather, I found it even without looking, simply by reflecting in music what your poems stirred up in me. People who know my music can bear witness to the fact that my first attempts to compose settings for your poems contain more of what has subsequently developed in my work  than there is in many a much later composition.

SCHOENBERG 1964, 35

In another letter, written a year later, Schoenberg reiterated Dehmel's influence on the younger generation: "Far more than any musical model, it was you who determined the platform of our musical experiments" (Birke 1958, 285).^[8] These are strong sentiments, but they constitute no great exaggeration when we consider the enormous role played by Dehmel in Schoenberg's early works: between 1897 and 1906 he was to sketch or complete at least eighteen works based on Dehmel's poetry. Among these are some of his most important early compositions.

The two earliest Dehmel songs, completed in September 1897, were Mädchen-frühling and Nicht doch! After a hiatus, Schoenberg returned to Dehmel in the spring of 1899 (as we shall see in the next two chapters). A single Dehmel collection, Weib und Welt (1896), was to provide the source for all of Schoenberg's Dehmel works of 1899. For the two settings of 1897, Schoenberg turned to an earlier volume, Aber die Liebe (1893). It seems significant that Schoenberg selected Dehmel poems that, although tinged with erotic subjectivity, nevertheless show their allegiance to the folk-like tradition represented by the Heyse texts for Mädchenlied and Waldesnacht. There is, however, enough difference in the poetry to

[7] See Sichardt 1990. After Schoenberg, Strauss composed the most Dehmel works. His ten songs, composed between 1895 and 1902, comprise Stiller Gang, op. 31, no. 4; Mein Auge, op. 37, no. 4; Leises Lied, Der Arbeitsmann, Befreit, and Lied an meinen Sohn, op. 39, nos. 1, 3, 4, and 5; Wiegenlied and Am Ufer, op. 41, nos. 1 and 3; Notturno, op. 44, no. 1; and Waldseligkeit and Wiegenliedchen. op. 49, nos. 1 and 3. None of these texts overlaps with the ones set by Schoenberg. On the personal and professional relationship between Strauss and Dehmel, see Schuh 1982, 437–45.

[8] For more extensive citation of these letters and discussion of their aesthetic implications for Schoenberg's style, see Frisch 1986, 138–42.

---

have liberated Schoenberg from the more obvious kind of Brahmsian symptoms and to have encouraged him to reach for something more individual, more "subjective."c

Mädchenfrühling

In Dehmel's poem, the arrival of spring is developed as a metaphor for the emergence of feelings of love and sexuality in the young maiden. The work derives its tension from setting the process of nature—that of April buds becoming May blossoms—against the unnatural recalcitrance of the boy, who seems not to share the vernal urges. In the first stanza the boy's failure to react is conveyed by the blunt, intentionally wooden rhyme of the oneand two-syllable lines, "und/s`ein Mund." In the last line of the poem the maiden's frustration is captured by the repeated "fühlt":

       Dehmel                                    Schoenberg

  1   Aprilwind;                               Aprilwind;
  2   alle Knospen sind                  alleKnospen sind
  3   schon aufgesprossen,           schon aufgegangen,
  4   es spriet der Grund,                es spriet das Grn,
  5   und                                            und
  6   sein  Mund                                 sein  Mund
  7   bleibr so verschlossen ?        bleibt so verschlossen
  8   Maisomenregen;                     Maisomenregen;
  9   alle Blumen langen,                 alle Blumen langen,
10   stille aufgegangen                   stille aufgegangen
11   dem Licht entgegen,               [line omitted]
12   dem lieben Licht.                     dem lieben Licht.
13   Fühlt, fühlt er's nicht?             Fühlt, fühlt er's es nicht [?]

      April wind; all the buds have burst forth [Schoenberg: opened], the ground
      [Schoenberg: green] is sprouting and  his  mouth remains closed?

      May sunshowers; all the flowers yearn, having quietly risen, toward the
      light, to the lovely light. Doesn't, doesn't, he feel it?

To the right of Dehmel's poem I give the text as it appears in both of Schoenberg's autographs.^[9] Two of the variants, the substitution of "aufgegangen" for

[9] Schoenberg's setting has been discussed in three studies: Bailey 1979, 82–85; Thieme 1979, 157–59; and Düling 1981, 149–54. Bailey and Dümling erroneously print as Dehmel's "original" the rather different, revised version of the poem that Dehmel prepared for his ten-volume Gesammelte Werke (Berlin: Fischer, 1906–9), where it appears in the collection Erlösungen. This situation is unfortunately typical of the Schoenberg literature, which often fails to take account of the complex publication history of Dehmel's poetry. As in the case of "Mädchenfrühling," Dehmel often transferred and revised poems between collections, both in the original individual volumes and in the two later editions of his collected works.

Schoenberg's personal library, preserved at the Arnold Schoenberg Institute, contains only the tenvolume Gesammelte Werke (and the individual volume Schöne wilde Welt [Berlin: Fischer, 1913]), which obviously did not serve as the source for his settings of 1897–1905 (see Steuermann 1979, 215). At the Schoenberg Institute there is also a single undated leaf in Schoenberg's hand, headed "Gedichte von Richard Dehmel," onto which the composer copied six poems from various sources: "Bitte," "Liebe," "Gieb mir," "Geheimnis," "Klage," and"Mannesbangen." Only the last of these was set to music.

---

"aufgesprossen" in line 3, and "das Grün" for "der Grund" in line 4, upset the original rhyme scheme of stanza 1 ("-sprossen"/ "-schlossen" and "Grund"/ "Mund"). The first substitution repeats (actually anticipates) the participle in line 10 of the poem. Thus it seems clear that these are inadvertent errors made in copying the text. The other principal alteration, however, the omission of Dehmel's line 11, may have been a more conscious structural decision, as I suggest below.

The music of Mädchenfrühling (Appendix ex. H) shows clearly that what gripped Schoenberg was less the textual detail than the sonorous potential of the opening word/image, "Aprilwind" (and also of the initial word of stanza 2, "Maisonnenregen"). The image of "Aprilwind" gave rise to the swirling arpeggiated accompaniment figure and the half-diminished harmony that dominate the song—sweep through it, one might say. The

harmony functions ostensibly as ii⁷ in B minor, and thus a preor subdominant chord.^[10] But since a firm tonic is avoided for the entire first stanza—B minor is touched on only in mm. 17–18, where it serves as a passing harmony between A minor and C major—the half-diminished seventh remains tantalizingly ambiguous or vagrant. The whole first stanza seems suspended above this and the related subdominant sonority, E minor, to which the dominant, , always resolves deceptively (mm. 7–8, 9–10, 23–24). Even the final cadence of the song, to B major (mm. 45–48), is made not via the dominant, but directly from the same half-diminished seventh, presented successively in root position, first inversion, and second inversion.

The first stanza is, indeed, a remarkable study in harmonic Flüchtigkeit and understatement. Schoenberg avoids establishing not only the tonic, but also the conventional secondary key, the relative major. The

harmony with the strong A pedal in mm. 13–14 would seem to indicate a move toward III, but no cadence is forthcoming. The expected A-major chord (V of III) appears instead as A mi-

[10] In Schoenberg's Theory of Harmony (Schoenberg 1978), this chord (not named as a "half-diminished") is illustrated on pp. 231–33 (ex. 164 f, g, i, k, and ex. 165) as deriving from the minor subdominant. Schoenberg later examines its more radical possibilities as a "vagrant" chord (the "Tristan" chord) on pp. 255–56 (ex. 189). These compositional possibilites are exploited by Schoenberg in a Dehmel setting, Mannesbangen, to be discussed in chapter 4.

---

nor (mm. 15–16), and moves through a passing B minor, to the Neapolitan C major (m. 19). The structural function of the A minor has been to introduce the crucial Neapolitan pitch

. The Neapolitan is the strongest subor predominant sonority employed to this point; but it fails, like the others, to lead directly to a dominant; instead, in m. 21, it moves to its own dominant, G. Only in the piano interlude is the dominant at last reached. In m. 22, the Neapolitan is replaced by the diatonic II, , which now leads to a definitive half-cadence on in m. 25.

The fleeting harmonies of the first stanza give way to more stable ones in the second stanza. The dominant of the interlude resolves to a broad B major. A pedal point on B is sustained for ten measures and leads purposefully to a halfcadence on

at "dem lieben Licht" (mm. 37–38). The "Aprilwind" halfdiminished chord reappears in the following measure, now functioning clearly— as it did not earlier—as subdominant, resolving to V on the downbeat of m. 44. In the postlude the original evanescent harmonic language returns, as the dominant dissolves back into the "Aprilwind" chord.

The remarkable harmonic planning and control in Mädchenfrühling are complemented by—actually, inextricably linked to—Schoenberg's sophisticated treatment of meter and rhythm, and by the large-scale design of the vocal line. After the strong downbeat orientation of the accompaniment in the first five measures, the syncopations of the left hand in mm. 6–10 come into conflict with the righthand figuration. This rhythmic displacement is a corollary to the harmonic avoidance of a firm cadence, discussed above. The end of the second vocal phrase is marked not by a clear downbeat, but by further ambiguity, a Brahmsian hemiola that creates one large measure of

out of mm. 10–11. Schoenberg then further attenuates the subdominant chord on the downbeat ofm. 12 by tying the E over the bar line; the subsequent D-major and A-minor chords in the left hand of mm. 13–16 are also tied. The hemiola returns for the piano interlude in mm. 23–24, leading at last to rhythmic-metrical stability with the arrival of B major at "Maisonnenregen."

The very active accompaniment of Mädchenfrühling is a foil to the laconic, understated vocal part, whose unfolding provides one of the finest and most fluent examples of developing variation among the early works of Schoenberg. The initial B–E motive materializes out of the accompaniment figure simply by picking two chord tones from the half-diminished chord. This "Aprilwind" motive then becomes the source for almost all subsequent vocal material, at least in the first strophe. The next two phrases both fill in and expand the range of the initial motive: "Alle Knospen" extends the range upward to

, "es sprießt" downward to A.

The developing variation also involves a progressively greater emphasis on the

---

upper

in the vocal line. The first phrase goes up to E, the second up toon the weak beats of mm. 7 and 9. Thereceives more prominence in phrase 4, with the quarter note on "und"; the pitch still, however, falls on a weak beat. With the first syllable of "Maisonnenregen,"appears for the first time on a downbeat; the phrase as a whole traces a broad arpeggio down to the same note an octave lower. The phrase "dem lieben Licht" climbs up to, cadences on, and sustains the, which is thereby given the greatest emphasis so far. And yet the process is not over, for as the piano returns to the opening material, the vocal line (m. 41) reaches briefly, wistfully still one half-step beyond, to the highest note in the song, the that represents the top of the "Aprilwind" figuration. One could scarcely imagine a more elegant, effective conclusion.

Finally, a word about the formal structure of Mädchenfrühling and a speculation on why Schoenberg omitted line 11 of Dehmel's poem. Schoenberg understood that the two stanzas of the poem, although roughly parallel, could not sustain a strophic or modified strophic setting: the musical form needed to reflect the process of growth or evolution described in the text. And yet the final line—where the maiden laments the boy's unresponsiveness—undercuts the optimism of the second stanza. To capture this design Schoenberg creates an abbreviated ternary or ABA' form. Since the A' is brief (mm. 40–48, or only 40–43, not counting the piano postlude), Schoenberg may well have felt it was necessary to truncate the poetry of the B section ("Maisonnenregen") somewhat in order to achieve a rounded musical form that would not be grossly imbalanced. He thus dropped one of the two clauses involving "Licht," a tactic that destroys the rhyme "regen / entgegen," but not the actual syntactical sense of the stanza.

The musical transition back to A' is fashioned most subtly. At m. 37, underneath the rising half-cadence of "dem lieben Licht," the piano begins to return to the opening "Aprilwind" figuration. The actual return, however, comes two measures later on the last beat of m. 39, as the

at the top of the arpeggio is replaced by the familiar . This return is achieved with a sophistication worthy of, and perhaps inspired by, Brahms, who often ushers in the reprise gradually in his ternary songs, so that we cannot point to a specific moment where it begins.^[11] Also Brahmsian is the way Schoenberg momentarily obscures the bar line at the beginning of the piano postlude. The "Aprilwind" figuration appears first in the left hand, instead of the expected right hand, on the last beat of m. 43; then the right hand begins an overlapping statement on the subsequent downbeat—a metrical position it has not occupied before. This statement is truncated to allow the right hand to begin its next statement in the "proper" place, on the last beat

[11] See, for example, the song Die Kränze, op. 46, no. I, mm. 47–50, discussed in Frisch 1984,100–101.

---

of m. 44. As in many passages of Brahms, the metrical ambiguity across the bar line between mm. 43–44 is momentary, but telling, adding interest or "development" to a recapitulation.

Perhaps the most remarkable aspect of these Brahmsian techniques is how well they are integrated into the newer, more individual musical idiom of Mädchenfrühling. Gone are the Brahmsian "symptoms" still evident in Mädchenlied and Waldesnacht; but the fundamental principles of Brahms's harmony, phrase structure, and thematic variation, remain—and are indeed used with still greater elegance.

Nicht doch!

The sense of evanescence, even incompleteness, imparted by the truncated form and ambiguous harmonic close of Mädchenfrühling suggests that the song was not meant to stand alone. I believe it was conceived as part of a complementary pair with the other Dehmel setting of 1897, Nicht doch! This also survives in two manuscripts, a rough draft (incomplete) and a fair copy. The draft, which breaks off at the measure corresponding with m. 32 of the fair copy, forms part of the same bifolium as the dated draft of Mädchenfrühling In addition to their association on music paper, the songs are linked by certain poetic and musical features. Schoenberg encountered the two poems on successive pages of Dehmel's Aher die Liebe. Although Dehmel himself does not specifically pair the two poems, they can be read—as Schoenberg apparently did—as a kind of before-and-after of the maiden's encounter with the young man. In the first poem he is unresponsive, but in the second he at last reacts, sweeping the maiden off her feet with a skillful seduction:

Mädel, laß das Stricken—geh,
thu den Strumpf bei Seite heute;
das ist was für alte Leute,
für diejungen blüht der Klee!
Laß, mein Kind;
komm, mein Schätzchen!
siehst du nicht, der Abendwind
schäkert mit den Weidenkätzchen.

Mädel liebes, sieh doch nich
immer so bei Seite heute;
das ist was für alte Leute,
junge sehn sich ins Gesicht!
Komm, mein Kind,
sieh doch, Schätzchen:
über uns der Abendwind
schäkert mit den Weidenkätzchen . . .

---

Siehst du, Mädel, war's nicht nett
so an meiner Seite heute?
Das ist was für junge Leute,
alte gehn allein zu Bett!—
Was denn. Kind?
weinen, Schätzchen?
Nicht doch—sieh, der Abendwind
schäkert mit den Weidenkätzchen. . .

Maiden, leave off your mending—go; put the stocking aside today; that kind
of thing is for old people; for the young, the clover is blooming! Leave
it, my child; come, my treasure! Don't you see the evening breeze flirting
with the willow catkin?

Dear maiden, don't still avert your gaze today; that kind of thing is for old
people; the young look each other in the face! Come, my child; look here,
treasure: above us the evening breeze is flirting with the willow catkin. . .

Do you see, maiden, wasn't it nice to be by my side today? That kind of
thing is for young people; old folks go to bed alone!—Well now, child? Are
you crying, treasure? Come now—look, the evening breeze is flirting with
the willow catkin . . .

Schoenberg created several significant musical connections between the two songs (see Appendix ex. I for the opening of Nicht doch!). Despite somewhat different tempo/mood markings (the fair copy of Mädchenfrühng is marked "rasch, etwas flüchtig, durchwegs leise," Nicht doch! simply "leich"), both songs seem to demand a similar pace. Such at least is suggested by their strikingly similar metrical-rhythmic profiles. Both have an almost continuous flow of sixteenth notes, and although Mädchenfrühling is notated in

and Nicht doch! in , a listener without a score might well hear both songs in the same meter (probably ), at least initially.

The upward flourish that occupies the first three beats of Nicht doch! is separated distinctly from the second half of the measure by a vertical wedge, a subito piano, and a change in figuration. It thus tends to be heard as a single measure of

. Indeed, in m. 6 Schoenberg treats this figure in precisely that manner, as an "extra" half-measure. The effect of m. i is then repeated in m. 9. In both songs the notated meter is overridden at the ends of phrases by an almost identical hemiola of in the accompaniment.

The songs display a similar piano figuration, especially at the opening. Although rising arpeggios or scales are, of course, common accompanimental devices in nineteenth-century Lieder, Schoenberg treats them in a special way here. If the two songs are played in succession, the opening of Nicht doch! seems to recall, and then suddenly abandon, the figuration of Mädchenfrüling. The upward-moving figure, whose contour recalls the "Aprilwind" of the first song,

---

Example 3.5
Nicht doch!  mm. 25–30 (piano part only).

gives way rather abruptly to a very different figuration oscillating between two notes a third apart. It is as if the upward-moving sixteenth-note figure has come to represent the kind of innocent, vernal love depicted in Mädchenfrühing, which gives way to the coarser, simpler sentiments of Nicht doch!

Mädchenfrühing and Nicht doch! also share certain tonal features, including a key signature of two sharps (also shared by the D-Major Quartet and Mädchenlied of the same year). Mädchenfrühiing is in B minor-major, Nicht doch! in D major. Although in the abstract this relationship could hardly be taken as evidence of harmonic planning, Schoenberg's treatment of D in the first song is suggestive. The D major sought but not reached in mm. 13–14 of Mädchenfrühiing (as discussed above) becomes the tonic of Nicht doch! —a relationship that becomes apparent only if the songs are performed together.

Although Nicht doch! is not as harmonically elegant as its companion song, one particular procedure points clearly to Schoenberg's growing control of large-scale tonal structure, control of a kind already discussed in the case of the contemporaneous D-Major String Quartet. The piano interlude that appears between strophes 1 and 2 (mm. 25–30; see ex. 3.5), and 2 and 3 (mm. 54–59), makes a sudden detour from the dominant seventh of D to the key of

, then after three measures comes around again, via G minor, to the dominant ninth that opens the strophe.

During the first two strophes this harmonic diversion to the flatted sixth, the upper neighbor of the dominant, remains a detail confined to the interludes. But in the third strophe, Schoenberg picks up and expands the idea. Here the second quatrain ("Was denn. Kind?") begins not in the tonic, as in strophes 1 and 2, but

---

in iii (Appendix ex. J).

major arrives with dominant preparation on the downbeat of m. 75 to accompany the appearance of the title words, "Nicht doch!" The harmony now works its way definitively back to the song's dominant by circumscribing the A in the bass: , arriving on the tonic in m. 77. The whole harmonic process is reiterated as the first two lines of the second quatrain are repeated (mm. 79–83, as happened in previous strophes), except that the relation of the vocal part and accompaniment is adjusted so that the arrival of the tonic now coincides with the words "Nicht doch!" As in the D-Major Quartet, then, Schoenberg has effectively integrated the flatted sixth degree and the tonic.

The two Dehmel songs of 1897 make a telling complement or sequel to the pair of Heyse settings. If the latter pair represent a saturation point of Schoenberg's Brahms absorption, the former two, especially Mädchenfrühiing, represent something of a new beginning. Had Schoenberg's encounter with the poetry of Dehmel been limited to these two works, we might not be inclined to see the poetry itself as initiating this new direction. But after a relatively fallow year in 1898, Schoenberg returned to Dehmel in 1899. The works of that year mark the real milestone in his early compositional development. This fact suggests that Dehmel's verse played a crucial role, one that must be examined further in the next two chapters.

---