Chapter Ten—
The Second Chamber Symphony, op. 38, and the Second String Quartet, op. 10 (1906–1908)

After completing the First Chamber Symphony in July 1906, Schoenberg immediately set to work on another one; the first sketches for the Second Chamber Symphony were made in August. Yet he was to complete neither this nor any other large-scale work for two years. From the winter of 1906-7 on, Schoenberg's writing of op. 38 (I shall refer to the Second Chamber Symphony by the opus number it eventually bore) was essentially overtaken by the composition of other works, including the songs opp. 12 and 14; the chorus Friede auf Erden, op. 13; the Second Quartet, op. 10, begun on 9 March 1907; and some of the songs of the Hanging Gardens cycle, begun in March 1908. (In Sketchbook III, which contains material for most of these works, there are also sketches from this period for songs based on the poetry of Dehmel, Goethe, Gottfried Keller, C. F. Meyer, and Hermann Löns, as well as for an opera based on Gerhard Hauptmann's Und Pippa tanzt.) Schoenberg returned intermittently to the Second Chamber Symphony, most notably during 1911 and 1916, but was not to complete the work until 1939.

It is not possible to give consideration here to all the significant works composed-let alone those sketched-during the extraordinarily creative two-year period between mid 1906 and mid 1908. This was a critical time in Schoenberg's early career, during which he began to make the break with triadic, functional tonality as an organizing principle, completed with the Piano Pieces, op. 11, of 1909. A detailed investigation of the technical means by which Schoenberg's music evolved into atonality (a word he himself disliked) is beyond the scope of this study.^[1] Here I shall attempt to convey something of this development through an

analysis of the two instrumental works that span this period, the Second Chamber Symphony and the Second String Quartet (the latter also, of course, a vocal work). The first movement of op. 38 can be seen as perhaps Schoenberg's last fully tonal masterpiece of the early period. The quartet begins to call into question the kind of procedures handled with such assurance in op. 38; indeed, the quartet is, in a sense, "about" the conflict or differences between tonal and atonal procedures.

An essential kinship between these two pieces has gone largely unnoticed in the Schoenberg literature, perhaps because the Chamber Symphony was completed only in 1939 and is thus taken to be one of Schoenberg's later works. But as Sketchbook III reveals, the two pieces were to a large extent composed in tandem during 1907-8 (see table 13). By the fall of 1907, simultaneously with the completion of the first movement of op. 10, Schoenberg had completed in the sketch-book a continuous draft, in short score, of almost all of the first movement of the Second Chamber Symphony (through m. 143) and 85 full measures of the second. He also began to copy a full score, which was then broken off in August 1908, shortly after the completion of op. 10. As far as it extends, this full-score fragment, printed in SW B11/1: 147-97, is in all important respects identical to the final version published as op. 38, apart from some aspects of instrumentation and orchestration. Thus we can legitimately assess the first movement up to the coda (m. 143) as a work of 1906-8.

It is not only the chronological proximity and the intertwined sketching that suggest that opp. 10 and 38 be considered together. They are also both Gegenstücke to their immediate predecessors in their respective genres: that is, the Second Chamber Symphony is to the First somewhat as the Second Quartet is to the First Quartet. This relationship is most obvious in the larger formal design. Both the Second Chamber Symphony and Second Quartet appear to have been conceived at the outset as (or, at any rate, soon evolved into) works in separate movements. In this sense they both depart from the larger four-in-one-movement design of opp. 7 and 9. The similarity extends further into the very stuff of musical style and compositional technique. Both op. 10 and op. 38 tend to move away from the dense, almost self-consciously dissonant and gritty contrapuntal style so characteristic of opp. 7 and 9. This is not to say that there is any shortage of counterpoint, but the overall impression, especially in the first movements, is of a more lyrical idiom and a less cluttered texture.

The sketches bear out this assertion to a large extent, especially if one compares the first notated thoughts for the respective opening themes of opp. 7 and 10. The sketch for op. 7 (SW B20: 36) shows initially only the bass and theme (some inner parts are filled in at mm. 6 and 9); it is essentially a two-part contrapuntal framework. The first sketch for op. 10 (SW B20: 174) is more clearly homophonic: a melody line supported by harmony (with some contrapuntal move-

ment indicated in the implied second violin line). (A similar harmonic conception dominates the sketches for the themes from mm. 16 and 43; see S₄ and S₅ .)

The effect of this more homophonic texture in both op. 10 and op. 38 is to place more emphasis—at least in the opening themes—on fluent motivic development in the primary voice. Both themes unfold by a process in which a very brief initial motivic gesture is presented, then immediately modified. In op. 38 (Appendix ex. R), the motive becomes displaced rhythmically by an eighth beat and takes on an additional eighth note; the melodic pattern is modified to move upward past the initial   to  . In op. 10 (Appendix ex. S), the one-measure motive becomes compressed rather than expanded, so that the high   in the varied repetition in m. 2 is now reached on beat 3 rather than beat 4. The rhythmic diminution allows for the addition to the melody of the B.

The Second Chamber Symphony, op. 38

Christian M. Schmidt has suggested that op. 38 represents a "regression" in Schoenberg's development during the years 1906–8:

Op. 9 opened up new paths: its harmony, with its extremely tight-knit relation to the motivic-thematic occurrences, thrusts out to the very limits of the tonal system. In this respect the Second Chamber Symphony represents a regression: neither can its harmony be regarded as a further step towards the dissolution of tonality, nor are its harmonic formations so organically rooted in the structure of the musical substance as is the case in Op. 9.

PREFACE TO PHILHARMONIA EDITION NO. 461

The practice of evaluating a work primarily on the basis of its progressiveness, although widespread in critical literature about the arts (and evident in portions of this study as well), seems especially wrongheaded here. The harmonic-motivic language of the work is certainly different from that of op. 9, but it is no less progressive. Indeed, op. 38 can be seen to represent an elegant refinement of some of the compositional techniques of op. 9.

The formal plan of the first movement is, to be sure, outwardly less adventurous than that of the double exposition of the first part of op. 9. Schoenberg turns to a large ternary design:

A (mm. 1-52)

B (mm. 53-94)

A' (mm. 95-140)

Coda (mm. 141-65)

The conservative aspect of this plan, however, disguises a musical structure and language of great fluency. By comparison with the first movements of opp. 7 and 9, the A and B sections of op. 38 are tonally more closed, in   minor and   minor respectively. The A' reprise is greatly transformed; segments of A are not only varied, but some are also transposed in a way that suggests that Schoenberg was still thinking, however subliminally, of sonata form. Although the A section certainly does not have the tonal thrust of a sonata-form exposition, its thematic process imparts an exposition-like feel. The design of the A section can be represented as:

A₁ (mm. 1-11), tonally ambiguous, moves to V of   minor

A₂ (mm. 11-22), begins in   minor, moves to V

A₃ (mm. 23-31), begins in   minor

A¹ ' (mm. 32-35), begins in   minor

A² ' (mm. 36-47), begins in   minor

Codetta, based on A₁ (mm. 48-53)

Although it provides the principal motivic-thematic material for the movement, A₁ almost has the feel of a slow introduction. It entirely avoids presenting the tonic   minor until m. 11, as Schoenberg himself proudly pointed out in an annotation on the full-score draft (SW B₁₁ /II: 92). And it unfolds almost tentatively, questioningly. Steady rhythmic motion comes only with the cadence to   minor in m. 11. In a sonata-form sense, then, A₂ might be said to represent the "first theme"; it is followed by the imitative A₃ , which might be said to serve as a "second theme." This in turn is followed by a substantial developmental passage of sixteen measures (A_1' -A_2' ), of the kind that often follows a second group in a sonata form.

Sonata-like procedures tend to be vestigial in the face of the wonderfully fluid thematic process that unfolds across the entire A section (Appendix ex. R). Indeed, Schoenberg may have consciously renounced overt sonata-like processes in order to focus on less "formal" thematic ones. It has been shown by Klaus Velten that the first movement of op. 38 is a supreme example of the simultaneous use of the Schoenbergian concepts of developing variation, in which thematic materials are generated from the continuous modification of a very few motivic ideas, and fluctuating (schwebende) tonality, in which many key areas are touched upon (Velten 1976, 91-98). Velten suggests that the themes tend to develop successively from the intervals of the fifth and half-step presented in the initial motive. "While the fifth leap stresses the tonic and thereby suggests the tonality, the chromatic step oversteps the tonal boundaries," he notes (ibid., 91).

Although he tends to remain only on the level of the individual theme and does not in fact take much account of harmony in his analyses, Velten is certainly right. Throughout themes A₁ and A₂ , one can hear Schoenberg quite consciously (even self-consciously) reworking, reshaping, and recombining the basic intervals of the fifth and the half-step (which we may call x and y). A₁ consists of six phrases or units leading toward a cadence in   minor in m. 11. These units cannot be said to shape themselves into any traditional antecedent-consequent or period structure. Rather they seem to unfold by a successive reinterpretation of the basic harmonic-motivic material. The first four units form complementary pairs (a, a', b, b') in which the first element of the pair presents a motive and the second modifies it in some way such as to lead toward a new harmonic area. The fifth and six (c, c') serve to "liquidate" the material and move toward the cadence. But even the use of the different letters a, b, and c tends to conceal the genuine continuity of the process.

We can hear quite clearly how b modifies the basic elements of a. Across mm. 3-4, x is contracted to a fourth,  , then followed by y. The process of contraction can be perceived as occurring through an inversion of y as the descending half-step  .  Indeed, the figure  , in which the inversion of y is attached to the beginning of a contracted (or inverted) x, becomes a significant enough motivic element that it can be given its own designation, z. At the end of b', the second interval of z is now expanded to a minor sixth. Phrase c consists solely of a presentation of z, in which the original descending fifth of x is restored. The original x is then repeated at the beginning of c'.

The harmonic component of the first movement of op. 38 is just as remarkable as the thematic-motivic one. The four phrases of theme A₁ move through a wide range of root relationships before reaching the tonic   minor. Some of these relationships seem to share in or reflect in particular the motivic identity of y, the half-step. Indeed, across the phrases a, a', b, and b', the roots rise by half-step from   to  . Phrase a moves from   minor up a half-step to A minor; phrase a' carries the same succession up a further half-step to a half-diminished seventh chord on  . Phrase b moves back to A minor; b' then pushes still further up to B minor. Phrase c and the first three notes of c' remain in the orbit of B, with moves to   minor and D major.

D major represents one of the remotest possible keys from the tonic: it is the key of the leading tone. The way in which Schoenberg moves from here to the tonic is fully characteristic of the range of his powers at this time. In m. 9, the   of the D-major chord is respelled as  . Underneath the sustained   comes the same   half-diminished seventh chord heard in phrase a' (m. 3), together with the original descending sixth motive   in the cellos and double bass. The melodic   is thereby redefined as a suspension within that chord. In another harmonic surprise, the   of m. 10 resolves not to the expected  , but to

 ; and with the concomitant change of   to   the half-diminished chord has reshaped itself as a dominant seventh of   minor. This change is indeed remarkable-almost magical-for half-diminished chords and dominant sevenths normally have very different tonal functions. That through voice-leading Schoenberg can transform one into the other shows both how flexible and at the same time how root-oriented his harmonic language is: the common root   is enough to give identity and logic to the chord during its evolution from half-diminished to dominant seventh.

The whole eleven-measure process of A¹ thus integrates extremely remote harmonies into what comes to seem a coherent, purposeful succession from the initial subdominant heard in mm. 1-4 to the dominant reached in m. 10. The harmonic process of theme A³ is also worthy of analysis. The first entry of the theme, beginning on the note  , functions essentially as a V-i cadence in   minor, sustaining the dominant adumbrated in m. 22. The tonic is reached on the second beat of m. 25. The second entry, beginning a half-step higher on   (here is another half-step relationship), fulfills an entirely different tonal function. Although the theme itself is transposed up a half-step through the  on the downbeat of m. 27, the second entry does not function as a dominant; instead it remains in B minor. By analogy to the first entry, the chord on the final eighth note of m. 26 should be a dominant of E minor. Instead, we get the dominant of B minor, which resolves onto B minor on the downbeat of m. 27. Schoenberg has thus radically reharmonized the three-measure theme while keeping its melodic component essentially intact. The third entry, beginning on G, is quickly followed by a stretto (new entries in mm. 28 and 29) and leads into the "developmental" return of A¹ in m. 32.

The manipulation of phrase structure and the harmonic vocabulary of op. 38 are different in significant respects from opp. 7 and 9, the two preceding instrumental works of the 1904-6 period. Perhaps the style of the first movement of op. 38 comes closest to that of the slower, more lyrical portions of these earlier works, especially the A-major theme of op. 9 (theme 11/2, m. 84) and the slow movement theme of op. 7 (K). But there is a distinct difference. The theme in op. 9 remains clearly rooted in the key of A major, which is articulated in mm. 86 and 91-92. The rate of harmonic change is slower; there are fewer different chords and the melody is thus more consistently and pungently dissonant with the underlying harmony. This intense kind of dissonant writing is part of the distinctive style of op. 9 (and op. 7). In op. 38, the simultaneous motion of the voices makes for more frequent local shifts of harmony; there are thus fewer dissonant melodic appoggiaturas.^[2] Another difference from op. 9 is in the nature of the harmonies. Theme 11/2 of op. 9 makes considerable use of dominant-seventh

[Full Size]

Example 10.1
Second Chamber Symphony, op. 38, 1, fourth chords at climax of A² .

chords not directly related to the local tonic A: specifically the D⁷ and   chords in mm. 85, 87, and 90. These kinds of chords, which help give the theme its peculiar flavor, are not present in op. 38.

A revealing point of comparison between the First and Second Chamber Symphonies is their respective treatment of quartal elements. As we recall, fourths play a prominent role in op. 9, especially at important formal junctures. We may recall too that the quartal harmonies of op. 9 tend to arise through stepwise voice-leading and that Schoenberg often seems to conceive these outwardly symmetrical formations in terms of root functions. There are no six-part quartal chords in op. 38 of the kind found in op. 9; from this point of view, the work is indeed less radical-more "regressive," in Schmidt's terms. But there are distinctly quartal elements, and the way in which they are integrated into the overall harmonic language is just as persuasive as in op. 9.

The first quartal sonority comes in m. 9, in phrase c' at the approach to the first cadence in   minor (Appendix ex. R). Underneath the sustained   of the flute, the violas play   and  . But any true quartal property of the chord is undermined by the root,   in the bass, which clarifies the chord (at least by the end of the following measure) as a dominant. Fourth chords appear more prominently at the climax of the A² ' theme, at m. 43 (ex. 10.1). In m. 43, the fourths are projected horizontally, in the melody. Genuine four-part fourth chords appear on the strong beats, alternating with normal triads (  and C respectively). The harmonic-melodic progression is carried along by strong stepwise voice-leading in the bass. This passage forms, on a smaller scale, a kind of structural analogy to that at mm. 355-67 in the development section of the First Chamber Symphony. There the treatment of the quartal sonorities reaches its climax in both the vertical and horizontal dimensions: both harmony and melody consist only of simultaneous and/or successive fourths. In mm. 43-44 of the first movement of the Second Chamber Symphony, Schoenberg achieves something similar. Yet, befitting the whole tone and style of op. 38, the climax is much more understated than in op. 9, although no less subtle.

[Full Size]

Example 10.2
Second Chamber Symphony, op. 38, 1, fourth chords.

Purer fourth chords are encountered at the important cadential points at the end of A and the coda, in particular at mm. 47-48 and 159-63. In the former (ex. 10.2a), the tonic is approached directly from a four-part fourth chord, which stands in for the conventional dominant. As in the climax four measures earlier, the radical nature of the cadence is muted, understated by the smooth voice-leading. The cadential passage at the end of the coda (ex. 10.2b), although apparently not composed or drafted in 1906-8, merits consideration here. The melodic figures of descending fourths are harmonized first by A minor, then by a four-part fourth chord built up from  , then a five-part fourth chord on  . Schoenberg moves back down through   with an exact transposition of the preceding five-part chord down a half-step, to A minor and back up to  . All these

[Full Size]

Example 10.3
Second Chamber Symphony, op. 38, 1, recapitulation of A³ .

chords are connected clearly and conspicuously by stepwise voice-leading. Schoenberg thereby manages to achieve logical yet untraditional continuity between sonorities and also manages, as in mm. 47-48, to avoid utterly any conventional V-i approach to the tonic.

This process of voice-leading is so smooth and so idiomatic that it forces us to redefine our own normal perceptions of tonal functions. Thus, it is hard to say whether the chord in m. 163 is to be taken as a substitute dominant, because of the   in the bass, or as a genuine tonic with the fifth in the bass (the tonic   is provided in the bass only at the very last moment, in m. 165). The simultaneous arrival of phrase a of the A¹ theme with the chord on the downbeat of m. 163 might argue for a "real" tonic. In any case, the very ambiguity shows how fluent Schoenberg's utilization of the quartal chords is-and how closely it is tied in with the half-step motivic figure y. The final cadence across mm. 164-65 contains, and in an important sense resolves, the   semitone relationship.

Before leaving the Second Chamber Symphony, we should devote some attention to the transformations of the A section in the reprise A'. It is here that we see in especially vivid form the synthesis of sonata and ternary procedures. The moment of return in m. 95 arrives quietly, unobtrusively. The original theme, accompanied by chords at its first appearance, is now surrounded by, or enmeshed in, a contrapuntal web comprised of three other voices: a syncopated theme in the first flute; a line descending in semitones, played by the bass clarinet and the first cello; and a theme played by the viola. The viola line is based rhythmically on the second, third, and fourth measures of theme A² (cf. mm. 12-14).

Real transposition occurs with the approach to theme A³ (ex. 10.3). In true sonata-like fashion, the theme begins (in m. 122) on  , down a fifth from its original appearance (m. 23). This brings the first entry to cadence on   minor, rather than the original   minor of m. 25. The second entry begins on  , also down a fifth from the original B, but Schoenberg modifies this entry so that its

final notes lead back to   minor. In its original appearance (Appendix ex. R, mm. 25-27), the second entry of the theme remained in B minor (unlike the first entry, which cadenced from V to i in   minor). Were the reprise exact, the theme would now sustain E minor: the last two notes (on the downbeat of m. 124) would be   (or   Instead, Schoenberg changes these to   down a half-step from the expected location. This change, together with the accompanying harmonies, serves to steer the theme back toward the tonic   minor. Such far-reaching transformations of A material in the reprise of op. 38 show Schoenberg manipulating sonata-like principles with extraordinary fluency.

The Second String Quartet, op. 10

As is suggested by Sketchbook III (table 13), Schoenberg began work on the Second Quartet in March 1907, about seven months after making the first sketches for op. 38. Over the next year and a half, the quartet was to supplant the Chamber Symphony as the focus of his creative activities. Although the genesis of the work as a whole was protracted, Schoenberg appears to have composed the individual movements at lightning speed; such, at least, is suggested by his claim to have written "three-fourths of both the second and fourth movements of my Second String Quartet in one-and-a-half days each" (Schoenberg 1975, 55).

Schoenberg himself saw the Second Quartet as marking "the transition to my second period" (Schoenberg 1975, 86) in two particular ways. First, large, continuous structures were replaced by separate movements (something also true, as I have suggested, of op. 38): "By abandoning the one-movement form and returning . . . to the organization of four movements, I became the first composer of the period to write short compositions. Soon thereafter I wrote in the extreme short forms" (ibid., 78). Second, the quartet intimated the renunciation of tonal centers that was to characterize the works of 1909: "In the first and second movements there are many sections in which the individual parts proceed regardless of whether or not their meeting results in codified harmonies. Still, here, and also in the third and fourth movements, the key is presented distinctly at all the main dividing-points of the formal organization" (ibid., 86). As has been suggested above (and will be elaborated below), Schoenberg's second point, about the juxtaposition of key-centered and non-tonal passages, forms the basic compositional premise of the quartet as a whole. In this important sense, the Second Quartet is different from either op. 7 or op. 9, pieces for which Schoenberg had made similar claims about the individual parts moving independently of harmonic implications.

As to the first point: although the division into distinct movements is, to be sure, significant, one of the great achievements of op. 10 is precisely Schoenberg's adaptation of the long-range cyclic or recapitulatory techniques of opp. 4,

5, 7, and 9 to the separate-movement format. One of the main themes of the first-movement exposition (1b, m. 12; Appendix ex. S) is recalled near the end of the trio in the scherzo movement (m. 180), just after the renowned "O du lieber Augustin" citation. The thematic material of the third movement, a setting of Stefan George's "Litanei" in variation form, comes almost exclusively from the preceding two movements; in this sense, the movement functions as a cyclic return. The coda of the finale is clearly intended to round off the quartet as a whole. The syncopated triplet figures (cello, m. 134) recall the opening theme of the first movement. The tonal return to the key of the first movement.   minor, followed by a resolution to   major, is also in this context, as Schoenberg himself acknowledged, a genuinely cyclic gesture.

The first movement of op. 10 is in outward respects a more conservative sonata form than the analogous movements of op. 7 or op. 9 (or op. 38). The exposition (mm. 1-89), development (90-145), recapitulation (146-201), and coda (202-33) are not hard to locate. (For Schoenberg's own analysis, which employs sonata form, see Rauchhaupt 1971, 43-44; see also the analysis in Whittall 1972, 20-21.) The exposition of op. 10 has the following design:

1a, m. 1

1b, m. 12

1a'/transition, m. 33

2a, m. 43

2b, m. 59

2c, m. 70

What is distinctive is the way Schoenberg manipulates the sonata structure into a technical and expressive conflict between tonal and atonal gestures, a conflict that prefigures that of the quartet as a whole. The contrast between themes, especially between 1a and 1b (and 1a and 2a, which is closely related to 1b, as will be shown below), is distinctly sharper than that between the analogous themes in op. 38 and is, indeed, perhaps more extreme than in any previous sonata form by Schoenberg.

Theme 1a (see Appendix ex. S) begins in   minor, passes through the chords of   major and A minor, and breaks off in m. 11 on a fortissimo F major. Even though the syntax or succession of chords is surprising and unconventional, the theme is clearly based on triadic structures; as such it seems almost to embody tonality or tonal processes. In theme 1b, no triad is evident; tonal associations are stretched to the limit through the primacy of counterpoint, dissonant appoggiaturas, and stepwise voice-leading. For the first five-measure phrase of 1b, the

melodic line alone could be heard as a conventional unit in , descending from scale degree 5 to a half-cadence on 7, a half-step below 1. A "normal" harmonization might move from 1 or i to V, perhaps with a strong G-major Neapolitan in m. 15. But the context into which Schoenberg places the phrase is far from harmonically normal. The melody is underpinned by two lines, one in the second violin, the other in the cello. The bass line descends in appoggiaturas by half-step from B to  . Except for the of m. 13, this descent fails utterly to support a progression from I to V. In this sense, then, we may speak of "atonality," fully realizing that the theme is not as resolutely atonal as the music written by Schoenberg after this quartet.

Schoenberg's bold stroke here-one that has implications not only for sonata form but for his entire musical style-is to reverse the expected associations of tonality/consonance with stability, and of atonality/dissonance with instability. Despite its triadic orientation, theme 1a is highly unstable. It refuses to hold to its   point of origin, but begins right away to wander tonally and in m. 11 literally falls apart on F major (one can certainly not speak of any cadence or close on F). The theme is also temporally unstable: it begins slowly, then speeds up dramatically, impulsively, across its eleven measures, as if propelling itself headlong toward its own demise.

From the viewpoint of phrase structure as well, 1a seems unstable. Outwardly, it has a two-part, antecedent-consequent structure of the kind Schoenberg himself might have analyzed as a "period" (mm. 1-7, 8-11). But the proportions and the general Stimmung are awry. The first phrase or antecedent begins with a proper two-measure unit, which is followed by a three-measure one. But the succeeding measures undermine any incipient regularity. The consequent is too short, and it is violent: instead of "completing" or complementing the antecedent, it grabs the theme and in m. 11 brings it abruptly to a halt.

The contrasting theme of the first group, 1b, presents a very different picture. It unfolds at a relatively constant tempo, which Schoenberg marks significantly as the Hauptzeitmaß, or principal tempo, almost as if what preceded is to be taken as Vorspiel. The smooth linearity of its bass line and the regularity of the melodic structure (5 + 5, mm. 12-16 and 17-21) give the theme a stability or solidity. Unlike theme 1a, 1b is accorded a full counterstatement beginning in m. 24. With its regular Hauptzeitmaß, its smooth voice-leading, and its substantial length, theme 1b, then, takes on the character of a principal or primary theme in a sonata form. In that a relatively unstable opening theme leads to a more stable second one, Schoenberg's procedure in op. 10 is analogous to that in the first movement of op. 38. But the tonal implications are now reversed. In op. 38, the reader will recall, the opening theme gives way to a firm (if temporary)   minor at m. 11. In the quartet, the triadic sonorities of 1a are made to sound less stable than the tonally more ambiguous 1b.

[Full Size]

Example 10.4
Second String Quartet, op. 10, 1, comparison of themes 1b and 2a.

The approach to the second group replicates (and thus reinforces) to some extent the process of the first twelve measures of the movement. The return of 1a as a transitional element (1a'/transition) brings with it the unsteady tempo of the opening measures. In a sense this passage shows 1a in its "true" light; it is more suited to be an unstable transition than an opening theme. The transition gives way in m. 43 to Zeitmaß (presumably the same as the Hauptzeitmaß of m. 12) and to theme 2a, which is derived directly and unmistakably from 1b.

This derivation or relationship becomes clear when the two themes are super-imposed as in ex. 10.4. Both consist of two phrases (5+5 measures in 1b, 4+5 in 2a). Both begin with a   neighbor motion and proceed with similar contours. It is perhaps not surprising to find that the two themes were actually sketched by Schoenberg very close together on the same page of Sketchbook III (see SW B20: 174-75). But more significant than their common or simultaneous genesis is their disposition in the sonata form. By using similar (indeed, aurally almost identical) themes in the middle of the first group and then at the beginning of the second group, Schoenberg is calling into question some of the perceptual foundations of sonata form. To be sure, Haydn had long before written so-called monothematic expositions, in which there is no distinct second subject to coincide with the arrival of the dominant. But in the later nineteenth century, when tonality (or at any rate, tonic-dominant polarity) no longer played such an important role in the articulation of sonata forms, composers tended to rely strongly on thematic dualism or contrast. As we have seen, Schoenberg certainly does so in earlier works, even where the second theme can be motivically derived from the primary one. In the first movement of op. 10, however, we seem to have a deliberate attempt to overturn the normal associations or conventions of thematic dualism.

Schoenberg also plays with the formal implications of thematic groupings. In one sense, the first 43 measures present an ABA' form, a conventional arrangement within sonata traditions from Schubert onward (see, for example, the first

movement of Brahms's Violin Sonata in G Major, op. 78, or of Schoenberg's own D-Major Quartet of 1897). But the ternary design, in which the longer, more stable theme normally encloses the shorter, unstable one, does not fit or suit the thematic material in op. 10. We are thus forced, or invited, to hear beyond the 43 measures to a larger two-part design, AB A'B', in which each part begins unstably and moves to a more stable theme (Appendix ex. S). The essential identity that Schoenberg creates between the B and B' themes (respectively 1b and 2a in my sonata-form analysis) overrides the conventional sonata plan, in which we expect a distinct contrast between first group and second group. In op. 10 Schoenberg has built that contrast right into the first group.

The continuation of the second group at m. 58 (not shown in ex. S) further complicates and extends the AB A'B' design by adding a suggestion of A": the theme that begins at m. 58 ("belebend"), which can be labeled functionally as 2b, or the second idea within the second group, is in fact derived rhythmically from the second measure of theme 1a (m. 2  of the movement).^[3] The derivation becomes explicit at the recapitulation, where 2b (m. 150) follows immediately upon 1a (m. 146) as part of the first group. The presence of 2b in the exposition as a developmental variant of 1a serves as a kind of complement to the relationship of 2a and 1b. In both cases, first-group material (1a and 1b) comes back more or less overtly in the second group (as 2b and 2a, respectively).

In the recapitulation, Schoenberg continues to explore the dualistic relationship between themes. At the opening of the recapitulation in m. 146, he strikingly reverses the process from the analogous part of the exposition. Theme 1a now begins in F major, the key or triad on which it fell apart upon its first appearance (m. 11); it then moves smoothly, through stepwise voice-leading to the tonic   minor in m. 159. Instead of the agitated, ever-accelerating tempo of the opening, the theme now begins slowly, broadly, and gets still slower across fourteen measures, so that the Zeitmaß of m. 159 really comes to seem quicker than the initial tempo. Schoenberg gives 1a still greater weight or status in the recapitulation by presenting the theme in augmentation in the cello (mm. 146-49). In these ways, then, the recapitulation of 1a effects a genuine reversal and accords theme 1a a stability it was denied in the exposition.

In his own analysis of op. 10, Schoenberg suggests that theme 2a is omitted from the recapitulation (Rauchhaupt 1971, 44). But it seems clear that the passage beginning at m. 196 serves precisely a recapitulatory function and leads to the coda at the resolution to   minor in m. 202. Furthermore, in following 1b (m. 159) and 1a'/transition (m. 186), 2a occupies the same position as in the exposition. At m. 196, theme 2a is actually divided into its two basic thematic units, which are presented simultaneously rather than successively. This is a typically

economical recapitulatory strategy for Schoenberg, one that is also present in the finale of op. 10. The half-step figure (from mm. 43-44, but without the whole-step descent of m. 44) appears in the upper voices; the descending figure (from mm. 45-46) is in the cello. The isolation of the half-step neighbor figure helps reinforce the close identity between themes 1b and 2a.

The first movement of op. 10 calls many sonata-like precepts into question, while nevertheless adhering with relative strictness to the formal prototype. The impetuous scherzo seems to throw any such caution to the wind. It is one of the most extraordinary and unorthodox instrumental movements among Schoenberg's early works, and certainly far more radical than the scherzos in opp. 7 and 9. Elaborating the Formübersicht prepared by Erwin Stein for the Philharmonia pocket score of op. 10 (no. 229), Schmidt has divided the scherzo proper into a brief, nineteen-measure "exposition" and a "development" lasting from m. 20 to m. 97 (SW B20: 178-79). Here the use of sonata-form terminology seems to me (even with my own strong propensity to read "sonata" into instrumental movements) misleading and inappropriate. To be sure, the absence of an immediate "recapitulation," or its displacement to the return of the scherzo after the trio, would perhaps not be surprising from the composer of the First Quartet and First Chamber Symphony. But in the Stein-Schmidt scheme, the proportion of "exposition" to "development" is out of balance. Moreover, that the three basic thematic ideas in the scherzo are separated by a fermata but no "transition" suggests a formal dynamic very different from sonata form. Indeed, the process seems almost to dismiss or turn its back on the rather orthodox sonata form of the preceding movement.

The body of the scherzo consists essentially of an alternation and cumulative development of three basic thematic units, A (mm. 1-13), B (14-17), and C (17-19), which seem constantly to react to each other (see Appendix ex. T). Unit A begins with only a rhythmic figure (x) in the cello (perhaps an echo of Beethoven's equally unusual scherzo in op. 59, no. 1?). In m. 4, this figure is overlaid with two other ideas (y and z). Beginning in m. 7, the upper two parts are inverted so that the staccato z appears below the legato y (whose rhythmic profile is now varied). In unit B, the rhythmic motive x, which was deprived of any melodic shape, is now accorded one; but after three and a half measures, C seems to dismiss the preceding lyricism with a chuckle of sixteenth notes. Later in the scherzo (m. 65), C is itself transformed from its original skittish shape into a lyrical melody "mit sehr zartem Ausdruck."

In the critical report to SW (B20: 178-89). Schmidt has described in illuminating detail Schoenberg's extensive sketches and drafts for the "development" section of the scherzo (mm. 20-97). His essential point is that the various thematic units that I have called A, B, and C (and that are called Gedanken 1, 2, and 3 by Schmidt), were continuously reshuffled and recombined, almost like pieces of a

[Full Size]

Example 10.5
Second String Quartet, op. 10, II, first sketch for scherzo theme, from Sketchbook III, p. 56.

puzzle, until the final form was reached (and indicated by Schoenberg in the sketchbook by a characteristic profusion of cross-reference symbols). The sketching process shows Schoenberg working in this movement toward a new kind, or a new degree, of developmental form.

It is worthy of notice too that according to the sequence of Sketchbook III, the scherzo was both the first movement of op. 10 to be sketched and the last to be completed (see table 13); it can thus be said to have preoccupied Schoenberg on some level throughout the eighteen-month genesis of the work (although he claims to have composed most of it in a day and a half). The first notated ideas for the quartet in Sketchbook III are contained in two sketches on p. 56 (see table 13; SW B20: 177). The first contains a single-line draft for the eighth-note motive z of theme A (ex. 10.5); it is identical to its final form except that it contains an immediate repetition of the first measure (and thus is three measures long in the sketch, rather than two as in the final version). The second appears to be a sketch for the trio and is, as Schmidt suggests, "very close to the 'O du lieber Augustin passage' in gesture and rhythmic shape" (SW B20: 177).

The first sketch is of relevance to the present study because it shows that Schoenberg's first notated idea for the Second Quartet was in the key of his previous quartet (and several other early works), D minor, and that, somewhat like the opening theme of op. 7, it seems intent on exploring the chromatic range around a D center. Indeed, this little theme can almost be taken as representative or symbolic of Schoenberg's tonal language at this time (probably early 1907). The tonic is obviously D, and it is complemented or supported by a clear dominant tone, A, as well as by two tonic-defining neighbor notes (part of the dominant chord),   and E. But the tonal field of D is, as it were, so sprinkled with chromatic "weeds" that the tonic becomes obscured. This overgrowth becomes especially evident when the theme is heard in its real context (as z), near the beginning of the scherzo. The D pedal sounding insistently in the cello throughout the first ten measures (x) is really tonic by assertion, above which y and z range widely over the chromatic space. In these ways, the scherzo of op. 10 can be considered the last great D-minor exploration of Schoenberg's early tonal period.

The disintegration of D becomes still more extreme in the trio, which is given a key signature of D major. Here the main theme, a combination of a rapid descending figure in the violin and a lyrical rising melody in the cello (ex. 10.6), is

[Full Size]

Example 10.6
Second String Quartet, op. 10, II, main theme of trio.

to my ear the least tonally centered of any so far in the quartet: that is, it has less audible relationship to D major than do the themes of the scherzo proper to D minor. Schoenberg refers to the violin figure of m. 98 as having "seven notes . . .  because this was the form in which this theme came to mind" (Rauchhaupt 1971, 45). Indeed, in the sketches for the trio, the figure consists entirely of eighth notes, divided between two measures,   (SW B20: 189). Schoenberg then changed it to a single measure of 4+3 "because I feared to be called a revolutionary." Not only the rhythmic profile, but also the tonal outline of this figure (which remained unchanged) might perplex a listener. The downward sequence through m. 101 descends by major third: the starting notes are respectively  , D,  ,  . Although this kind of symmetrical division of the octave was hardly new in 1907 (it already appears in Schubert almost a century earlier), its combination with a figure that is already highly chromatic and unstable makes for virtual atonality. In this sense, the sequence is not unlike the more famous and vivid one at the opening of the fourth movement.

The cello theme that is placed underneath the sequence beginning in m. 100 is also extremely dissonant. Its nine notes expose eight different chromatic tones (only the   appoggiatura in m. 103 is a repetition) and fail in any sense to project D major (or any other key). The pure A-major "dominant" triads that follow in mm. 104-5, 107, and 108 seem intended as a witty-but ineffective-corrective to the wanderings of the first theme.

The essential dramatic scenario of the scherzo and trio of op. 10 appears, then,

[Full Size]

Example 10.7
Second String Quartet, op. 10, II, transformation of  "Augustin" theme.

to be the dissolution or evaporation of conventional tonality. This tale reaches its overtly programmatic climax at the end of the trio, in m. 160, where the musical development of the trio themes screeches to a halt and yields to the renowned quotation of "O du lieber Augustin." The intrusion of this popular song, especially of its last phrase, "Alles ist hin," serves, as is often remarked, as a kind of self-referential commentary on the distintegration of the musical language. What is perhaps not so often articulated is what Schoenberg himself points out (Rauchhaupt 1971, 45): that the liquidation or taking-apart of the "Augustin" theme in mm. 171-92 yields the principal motives of themes 1b and 2a of the first movement. Specifically (ex. 10.7), the opening neighbor-note figure, A-B-A (or  ), of the song (m. 165) becomes transmuted first (m. 180) into the head motive of 2a ( ), then into the  figure of 1b (at m. 187). Underneath this latter transformation, Schoenberg turns the other part of the "Augustin" melody, the descending fifth (E-A, m. 167) and the rhythmic figure it embodies (three quarter notes with an accent on the first), into the third and fourth measures of theme 1b.

Here we have an extraordinary example of the kind of thematic transformation that we have been tracing throughout Schoenberg's early works and that is now put into the service of a programmatic statement. With the transformation of the banal, tonal street song back into the intense, dissonant thematic material of the quartet, Schoenberg seems to be saying that the step between tonality and atonality (or between consonance and dissonance), which was presented in exaggerated form in the opening segment of the first movement and then further exposed in the scherzo and trio, is in fact not so great. It is not an unbridgeable, absolute gap. This is, of course, the viewpoint repeatedly propounded by Schoenberg in his Theory of Harmony, where he argues that "dissonances are the more remote consonances of the overtone series" (Schoenberg 1978, 329).^[4] Here, in the trio of

op. 10, he makes the point musically by referring back to precisely that spot in the first movement where the tonal-atonal or consonant-dissonant juxtaposition was made, in the first theme group. It is to create this association for the listener, rather than to fufill any purely "cyclic" impulse, that Schoenberg here makes reference to a theme from the first movement. Thus in this way, as in so many others, the Second Quartet makes over traditional techniques.

The apparent genesis of the last two movements of op. 10 sheds further light on the important issue of cyclicism or return in the work. According to the sequence of sketches, as well as certain notations, in Sketchbook III, the present fourth movement, "Entrückung," was originally projected to be in third position. The sketches for it are headed "III Satz Streichquartett"; they follow sketches for movements 1 and 2, and precede those for "Litanei" (see table 13). "Litanei" was thus not only the last movement conceived (although the scherzo was the last actually to be completed), but appears to have been intended as the finale, even though no sketches are marked specifically as such. Such is the reasonable speculation of Schmidt, who notes that an ordering "Entrückung"- "Litanei" would have placed the most obviously recapitulatory movement last, according to nineteenth-century tradition (SW B20: 193).

I would concur with Schmidt's interpretation of the evidence, in part because, judging by the example of Schoenberg's earlier instrumental works that have been examined in this study, we would expect him instinctively to place some kind of important large-scale return near the end of a piece. The key of "Litanei," however, which is   minor even in the earliest sketches, presents something of a challenge to the theory that the movement might have been intended for last place. As we have seen, Schoenberg (unlike Mahler in this period) tends to end his large-scale instrumental compositions in the key in which they began. It is certainly possible that for op. 10, which departs from the earlier works in many respects, Schoenberg would have opted for a tonal conclusion away from   minor. But in a sketch for stanza 4 of "Litanei" there is strong evidence that he did indeed plan a return to   minor, one that would moreover coincide with a direct recall of theme 1a from the first movement, in its original triple meter ("Litanei" as we know it remains in   throughout). In this sketch (ex. 10.8; see SW B20: 194), in which a tonic and dominant of   minor are clearly articulated, the bass is ingeniously fashioned from what I called theme 2b of the first movement (m. 58), a theme that also plays a large role in "Litanei" (first appearing in augmentation in mm. 4-9). As was pointed out above in the discussion of the first movement, theme 2b was itself derived from 1a, and the two themes were juxtaposed in the recapitulation; their actual contrapuntal combination in the "Litanei" sketch thus seems especially appropriate. This sketch was rejected when "Litanei" assumed its third-movement position in the completed work.

When he reversed the order of "Entrückung" and "Litanei," thus displacing

[Full Size]

Example 10.8
Second String Quartet, op. 10, III, sketch for fourth stanza of  "Litanei," from Sketchbook III, p. 108.

the most substantial cyclic return from its customary end position, Schoenberg effected a profound change not only in the shape of the quartet but in his whole conception of large-scale form. From the viewpoint of the texts alone, the decision was both significant and logical. In the final version of the quartet, the questioning, anguished persona of "Litanei" is liberated and released in "Entrückung," literally set afloat in the atmosphere. Here the "qual" (torture, anguish) of "Litanei" is completely "erloschen" (extinguished). George uses the word qual in both poems, and Schoenberg clearly picks up on the verbal association. An ordering of movements "Entrückung"-"Litanei," as intially planned, would have given a completely contrary motion, from liberation toward pain and suffering (albeit with the hope of redemption).

As is well known, Schoenberg dedicated the Second Quartet to his wife, Mathilde. The composition of the work coincided with a particularly painful point in their marriage, just after the end of Mathilde's affair with the painter Richard Gerstl (who committed suicide). Schoenberg's apparent vacillation over the ordering of the last two movements may well reflect his own emotional conflict over the Gerstl affair. From an initial pessimistic despair, from an "Entrückung" - "Litanei" ordering, he moved toward a more optimistic, life-affirming (really art-affirming) position, signified in part by the ordering "Litanei"-"Entrückung."

Although many commentators, including Schoenberg, have a tendency to speak of "shorter" forms in op. 10, the finale is, in fact, an expansive structure. Lasting almost eleven minutes, it is twice as long as any of the preceding three movements and half as long as the entire First Chamber Symphony. In its fusion of song and instrumental form, "Entrückung" is one of the masterpieces of Schoenberg's early period. As in the first movement, and as in so many other structures we have examined in this study, Schoenberg relies on sonata-like pro-

cesses to give the finale its shape. In his Formübersicht in the Philharmonic pocket score, Erwin Stein gives the large outlines of the sonata structure, which may be expanded or filled out as follows:

INSTRUMENTAL INTRODUCTION

Part I, mm. 1-9

Part II, mm. 10-15

Part III, mm. 16-21 (= transition to exposition)

EXPOSITION

Theme 1, mm. 21-26

Transition/development, mm. 27-51

Theme 2, mm. 51-66

DEVELOPMENT

Part I, mm. 67-82

Part II, mm. 83-99

REPRISE

Themes 1 and 2 combined, mm. 100-19

CODA, mm. 120-56

The renowned introduction to this movement is a masterpiece of tone painting that both illustrates the "transport" of the poem and epitomizes or captures the essential tonal-atonal conflict that lies at the basis of the quartet. The opening of the introduction is built from essentially two different melodic figures or types. The first, introduced by the instruments in rising imitation in the first measure, can be called the "floating" figure. Although if the entries come at the interval of a fifth ( ), thus paying lip service to contrapuntal tonal tradition, the "subject" itself has no tonal base; it consists of eight different notes (categorized as pitch-class set 8-12 in Forte 1978, 165). The floating figure and its close relatives dominate in the upper strings in mm. 2-5. Underneath, in the viola and cello in m. 3, Schoenberg introduces the second basic motive, the descending fifth. With its downward, cadence-like gesture, this fifth works as a kind of counterforce to the floating figure and thus might be said to embody "gravity."

The descending fifth and the open-fifth interval itself play an important role in this movement. In the introduction, they serve not only as a melodic figure, but also as closing gesture for parts I and II. Part I ends with an emphatic pizzicato G-C (m. 9); this same fifth is reiterated as a simultaneity at the end of part II (m.

15). It is then raised a half-step to   to begin part III, which serves as the transition to the entry of the voice. In the primarily atonal context of this introduction, the prominent fifths cannot be said to be harmonically functional: that is, parts I and II cannot be said to end or cadence in C major. Nor can the successive entries of fifths in mm. 3 or 6 be said to establish any key (although as Forte rightly points out, the succession in the first violin in m. 6 produces the   scale, enharmonically  , the tonic of the movement [Forte 1978, 170]). Rather, to my ear, the fifths seem somehow to symbolize the tonal processes Schoenberg is about to abandon.^[5]

The tonic of this movement makes its first real appareance, over a dominant pedal, at the cadence in theme 1, "ich fühle luft." This theme, as suggested above, functions much like the first theme in a sonata form, essentially stable and tonally focused. It is followed by a developmental transition based entirely on motives derived from the introduction and from theme 1. In the tonal world of this movement, as in the first movement, we cannot expect the traditional harmonic markers of sonata form—that is, a clear modulation to a second key area within the exposition. Nevertheless, the entry of "ich löse mich in tönen" functions much like a second theme, except that it is still more strongly oriented around the tonic than theme 1.   now appears in root position, and the theme moves harmonically from I to V in mm. 54-55 (the dominant seventh chord has the fifth in the bass).

The development divides into essentially two sections, of which the second, beginning at m. 83, is remarkable for having the feel of a lilting scherzo or waltz. Here Schoenberg transforms the previously angular, staccato triplets into lyrical melodic lines. The floating of the introduction has turned into a more rhythmically defined dance on the "sea of crystalline radiance."

We have looked frequently in this study at the phenomenon of recapitulation in Schoenberg's early works and have investigated some of the ways in which he manages to effect a reprise of material without any exact repetition. In the first movement of op. 10, the reader will recall, Schoenberg compressed the reprise greatly, in part by presenting the two halves of theme 2a in counterpoint with each other. This kind of compression is carried even further in the finale. Here Schoenberg brings themes 1 and 2 together contrapuntally. The voice part at "ich

bin ein funke" is a reprise of theme 1, first heard at "ich fühle luft." Surrounding this theme above (first violin) and below, or roughly in the same register (second violin), is theme 2 in octaves. Also brought in from theme 2 here is the bass line ascending stepwise from the tonic  . Originally this bass had ascended only as far as   (m. 59). Now it continues up to the A a tenth above its starting point. Here the line breaks off and the bass drops down to a low D (m. 110).

Now begins perhaps the most astonishing aspect of the reprise: theme 1, at its original pitch level (D-G-A-C), appears in augmentation in the bass underneath the climactic final words "heilige stimme." A sequential repetition begins in m. 114 on  , but is broken off on the third note,  , with the end of the vocal line (m. 116). This reprise of theme 1 as bass is logical because at its first appearance in m. 21, the theme was the bass-that is, the lowest voice in the texture. When it appears at the beginning of the recapitulation, "ich bin ein funke," the theme is, as noted, accompanied by the stepwise bass brought in with theme 2, and it is accordingly transposed to begin in the key area of  . Thus Schoenberg felt it appropriate to bring theme 1 back once again at its original pitch level and in its original function, as bass. The vocal part at "heilige stimme" can also be heard as part of this recapitulatory process, since it begins like the viola counterpoint in the original "ich fühle luft" presentation:   (mm. 21-22). The condensed recapitulation of the finale of op. 10, lasting only nineteen measures, is surely one of the finest of Schoenberg's early period. It also serves as a harbinger of the kind of polyphonic compression, or simultaneous presentation of themes, that was to characterize his later music.

The moving instrumental coda to the movement serves to balance the introduction and, as has been suggested, to round off the entire quartet. If the introduction can be said to have stressed "floating" and atonality, the coda moves in the other direction, toward tonal resolution. Here we get the longest sustained tonic and dominant phrases in the movement. The final cadence of the work is, however, made not from the dominant, but from the trichord   that has appeared frequently throughout the movement: as the first chord accompanying theme 1 (m. 21); at the original-pitch reprise of theme 1 in m. 110; and again in the coda (m. 140, now as a D-minorish tetrachord with  ).

The cadential resolution is combined with or accompanied by a reference to the initial "floating" motive of the introduction, and as such it seems to epitomize the remarkable blend of traditional tonal procedures and newer atonal ones in this quartet. From the high  , the first violin makes a fifth descent to   The stark D-A fifth in the cello, the relative of the open fifths we have heard since the introduction, proceeds by parallel motion to the tonic fifth  . The A, already present in the trichord, remains as part of the tonic minor articulated in the penultimate measure, then moves stepwise to the major third  . It is hard to envision a more satisfying cadence; it is still harder to articulate why the conclusion

is so right. At this point, critical analysis must leave off and mute admiration take over.

In his writings Schoenberg attributes to his subconscious, the "Supreme Commander," many aspects of the composition of his works, such as elegant motivic and thematic relations. In the final stanza of "Entrückung," we encounter an artist who has in effect placed himself in the hands of, or made himself the vehicle of, a "heilige stimme" beyond his own direct, conscious control. It is clear from the music he wrote for this movement that the "stimme" was now leading Schoenberg away from many of the compositional preoccupations traced in this study. Consideration of op. 10, then, serves as an appropriate end point.