Uh-oh, it looks like your Internet Explorer is out of date.
For a better shopping experience, please upgrade now.
Arnold Schoenberg is widely regarded as one of the most significant and innovative composers of the twentieth century. It is commonly assumed that Schoenberg's music divides into three periods: tonal, atonal, and serial. It is also assumed that Schoenberg's atonal music made a revolutionary break with the past, particularly in terms of harmonic structure. This book challenges both these popular notions. Haimo argues that Schoenberg's 'atonal' music does not constitute a distinct unified period. He demonstrates that much of the music commonly described as 'atonal' did not make a complete break with prior practices, even in the harmonic realm, but instead transformed the past by a series of incremental changes. An important and influential contribution to the field, Haimo's findings help not only to re-evaluate Schoenberg, but also to re-date much of what has been defined as one of the most crucial turning points in music history.
About the Author
Ethan Haimo is a theorist and composer, and lectures in the Department of Music, University of Notre Dame.
Read an Excerpt
Cambridge University Press
9780521865425 - Schoenberg’s Transformation of Musical Language - by Ethan Haimo
1 “Atonality”: a revisionist thesis
… many people, instead of realizing its evolutionary element, called it a revolution.1
In this period I renounced a tonal centre – a procedure incorrectly called “atonality”.2
In 1939 a little-known author and poet, Ernest Vincent Wright, published a book entitled Gadsby.3 If Wright’s book were to be judged solely on its belletristic merits it would have slipped into obscurity long ago; rare is the book that is so utterly lacking in literary finesse.
If Youth, throughout all history, had had a champion to stand up for it; to show a doubting world that a child can think; and, possibly, do it practically, you wouldn’t constantly run across folks today who claim that “a child don’t know anything.” A child’s brain starts functioning at birth; and has, amongst its many infant convolutions, thousands of dormant atoms, into which God has put a mystic possibility for noticing an adult’s act, and figuring out its purport.4
The awkwardness of this passage is palpable. The odd turns of phrases, the stilted and pretentious vocabulary, the inexplicable use of an ungrammatical quotation, the rambling argument, and the generallyinefficient use of language all combine to create a very inept impression. There is, however, a reason for this ineptitude and it is for that reason that Wright’s book has not slipped into an otherwise well-deserved obscurity. As Wright proudly announced in the subtitle, his book was “A Story of Over 50,000 Words Without Using the Letter ‘E’.”
Why would an author undertake such a task? Why would he want to exclude the use of the definite article, the past tense of weak verbs, all third-person pronouns, not to mention an estimated forty percent of English vocabulary? As Wright explains in the introduction, he did so “not through any attempt to attain literary merit,” but “due to a somewhat balky nature, caused by hearing it so constantly claimed that ‘it can’t be done’. ”5
Mindful of, and presumably sobered by, this precedent, I nonetheless propose to talk about Schoenberg’s compositions from 1899 to 1909 and (other than in this and the final chapter) to avoid entirely the use of the word “atonal”. Unlike Wright’s Gadsby, however, I believe the reasons for this restriction are anything but gratuitous. Without fear of hyperbole I would even venture to state that a proper understanding of the music in question is not possible unless we can start to free ourselves from the word “atonal”.
To be sure, the case for using the word “atonal” cannot lightly be dismissed. It has come to be so widely accepted that it might appear to border on sheer contrariness to restrict its use. One might well be justified in wondering whether avoiding the word “atonal” would inevitably lead to awkward locutions like those that abound in Wright’s novel. Finally, even if one argues that the literal meaning of “atonal” is inaccurate or misleading, the way in which the word has been used for more than a half-century should have established a contextual meaning that can be understood independently of its literal meaning.
All of these arguments in favor of the use of “atonal” might seem persuasive enough, but I believe that they are more than outweighed by the counter arguments. For one thing, Schoenberg himself came to dislike the term. From the 1920s on, he used it sparingly and reluctantly, often with scare quotes. He resisted the term because he argued that it was illogical:
I find above all that the expression “atonal music”, is most unfortunate – it is on a par with calling flying “the art of not falling”, or swimming “the art of not drowning”. Only in the language of publicity is it thought adequate to emphasize in this way a negative quality of whatever is being advertised.6
Schoenberg went on to state that:
… this expression is wrong: with tones only what is tonal, in keeping with the nature of tones, can be produced; there must at least be that connection of tones based on the tonal, which has to exist between any two tones if they are to form a progression that is at all logical and comprehensible; an opposite, “atonal” can no more exist among tones and tone-relationships than can an opposite “aspectral” or “acomplementary”, among colours and progressions of colours.7
Schoenberg further claimed that the term was not chosen after a careful examination of the properties of the works in question. He stated that it originated as a term of scorn, coined by a journalist looking for a catchy phrase with which to pillory his music.
Moreover the expression, atonal cannot be taken seriously as an expression, since that was not how it first came about; a journalist derived it by analogy from amusisch, as a means of overaggressive characterization – such, at least, was the context in which I first noticed it.8
Finally, if he were to choose a single term to describe this repertoire, Schoenberg felt that something other than “atonal” might better capture the sense of the music. Accordingly he suggested “polytonal” or “pantonal”, terms that imply an expansion or evolution of past procedures, not their utter abrogation.9
A composer’s view of his music is important; it deserves respect and attention. But it is not necessarily the final word. Even though Schoenberg both rejected the term “atonal” and provided us with some alternatives, his preferences need not be binding. We have the right to overrule Schoenberg’s preferences if we can demonstrate that the term “atonal” really is the term that best characterizes the music in question.
I do not believe that such a case can be made. To the contrary, I think there are so many negative consequences that flow ineluctably from the use of the term “atonal” that we are far better off abjuring its use.
We can start by picking up where Schoenberg left off, with the recognition that atonality is a negative term. By its literal meaning, it seems to suggest that the one significant feature of an atonal composition is that it is not tonal. But precisely what does that mean? The closer one looks at this definition, the more dubious it becomes.
If atonality merely means that a composition lacks a tonic, then the definition is so broad as to be useless.10 Many different kinds of compositions – from the twentieth century and before – lack an identifiable tonic. If compositions as diverse as Stravinsky’s Rite of Spring, Lasso’s Prophetiae Sibyllarum, and Schoenberg’s Erwartung, Op. 17 are all “atonal”, then this definition has little value: it offers no means of discriminating between dramatically different compositions whose pitch languages have almost nothing in common.11
One does little better by stating that it is the absence of a concluding triad that makes a composition “atonal”. Here we have not only a negative definition but a highly reductive one as well. That Schoenberg’s post-1908 compositions conclude with something other than a triad is not sufficient reason to group them together into a coherent period or to distinguish them from their immediate predecessors.
We might try to refine the definition by asserting that “atonal” does not just mean lacking a tonic or a concluding triad, but rather denotes music which lacks all of the important structural features of common-practice tonal music. According to this revised definition, anything that tonality is, atonality is not. If tonal music has a tonic, then atonal music must not. If tonal music employs a referential diatonic collection, then atonal music cannot. If tonal music differentiates between consonances and dissonances, with the former resolving to the latter, then atonal music will not make such differentiations. If tonal music employs triads and classifiable seventh chords and organizes these chords into coherent harmonic progressions, then atonal music must not.
On the face of it, this definition seems more logical, and one could argue that there are a number of twentieth-century compositions that do conform to this definition. None of those, however, is to be found among Schoenberg’s pre-serial works (and perhaps not in the serial works either).12 The problem is, when we actually look at Schoenberg’s music from ca. 1909, we find that the revised definition is problematic precisely because it is inflexibly negative. Even in compositions where it is clear that there is no tonal center, where triads are not used as referential sonorities, and where the intervallic relationships are not in conformity with the traditional definitions of consonance and dissonance, there are still many features on the surface that bear a clear kinship with features we would find in Schoenberg’s earlier compositions. Triads and various kinds of seventh chords do make appearances as chords, or as parts of chords, or as the components of linear successions. Various scale segments still appear on occasion. Melodic lines often outline patterns that are similar to those we find in previous works. There are also rhythmic/melodic formulas that sound suspiciously like patterns from earlier music. So too, most of the techniques of voice leading are indistinguishable from those of earlier repertoires. And the consonance-dissonance distinction lives on, albeit significantly transformed.
If we are to take our revised definition of atonality seriously, these elements pose a problem. The revised definition is founded on the assumption that atonality must mean the absence of anything tonal. Since many such apparent “anachronisms” appear in Schoenberg’s post-1908 compositions, we are left with two options. Either we face up to the reality of the musical surface and entertain the possibility that these supposed anachronisms have structural significance (in which case the revised definition is useless), or we must deny, or at least downplay, the significance of these relics.
It is this latter attitude that underpins the analytical method most used for the explanation of the pitch structure of Schoenberg’s post-1908 compositions. It is fair to say that since the 1970s, pitch-class set analysis has come to be the dominant analytical technique for this repertoire.13 As it is typically practiced, pitch-class set analysis tends to downplay the significance of some of the prominent structural features of prior repertoires (triads, seventh chords, diatonic collections, and scalar segments) when they appear in Schoenberg’s “atonal” music. For example, regarding the opening motive in the vocal line in Schoenberg’s “Herzgewächse” Allen Forte has stated:
Similarly, the opening “motive” of the voice, F-G-G-F♯-F♯, is not in itself a structural set, but a secondary formation, a nonset, the elements of which belong to other sets. In general, a given “melodic line” may not necessarily be a discrete structural component. This facet of Schoenberg’s music has led to much misunderstanding and caused many blunders in the past, particularly where such secondary formations are “chromatic lines” (as in the case of the voice motive here), “whole-tone scales,” or other familiar patterns. Schoenberg simply did not compose with these and other well-worn formulae, just as he did not compose with sets that properly belong to the vocabulary of tonal music – although, as will be evident in subsequent examples, such sets may be indirect results of the interaction of significant structural components.14
In this book I argue that it is mistaken to minimize the central role played by the many “well-worn formulae” that lie directly on the surface of Schoenberg’s post-1908 music. The frequent presence in Schoenberg’s post-1908 music of “familiar patterns” and “sets that properly belong to the vocabulary of tonal music” cannot be dismissed as merely coincidental to more important structural features.
The term “atonality” is also problematic because it encourages us to group all compositions defined as “atonal” into a single class. Implicit in such a grouping is the assumption that all of the works of the class share their most essential procedures and techniques.
This too is misleading. That all of Schoenberg’s completed compositions between 1909 and 1920 lack triads as the concluding sonorities is not sufficient to support the thesis that they form a coherent period. Grouping all of these works into a single class promotes the mistaken notion that one aspect of the pitch language (the absence of concluding triads) is the only significant criterion for categorization. This is patently wrong. As we shall see in the last chapter of this book, there are far more meaningful and useful ways of dividing Schoenberg’s works into periods than whether they do, or do not, conclude with triads. To assert the existence of an “atonal” period is to lump together works that are radically different from one another.
There is yet another conceptual problem prompted by the use of the term “atonal”. Since Schoenberg’s post-1908 compositions have come to be referred to as “atonal”, it follows that the compositions before that point have been identified as “tonal”. Admittedly, the term “tonal” does not have the drawback of being a negative term, but here too the reality of Schoenberg’s compositions is rather more complicated than can be encapsulated in a convenient slogan. It is true that all of Schoenberg’s pre-1909 compositions do conclude with triads. As such, there is a natural tendency to think of these compositions in terms of theoretical paradigms that we employed for the enormous body of compositions that are included in the “tonal” category. To do so, however, is to operate from assumptions that may not be valid. Schoenberg’s pre-1909 compositions can no more be effectively described by a single slogan than can his post-1908 compositions. Schoenberg’s approach to tonality was a complicated and ever changing reality. As early as the songs of Opp. 2 and 3, his compositional approach included essential elements that are not readily compatible with the way tonality had been preached and practiced for more than two centuries in European classical music.15 Over the course of the subsequent decade, those radical elements continually increased in significance. To refer to all of his pre-1909 compositions as “tonal” is to gloss over any recognition of this extraordinary transformation.
As indicated above, one of the negative consequences of the word “atonality” has been that its use has helped lead to a failure to admit that significant elements of the past persist in Schoenberg’s “atonal” compositions. In embracing that reality we must not make a complementary error. That elements of his earlier pitch language persist after 1908 should not be taken to mean that Schoenberg’s post-1908 compositions are merely modified tonal compositions. In particular, I firmly believe that key centers simply do not play any kind of a functional role in these works. Elements of Schoenberg’s prior vocabulary and syntax remain in his post-1908 compositions; this does not mean these compositions merely present another way of expressing referential tonal centers.
There is a long tradition of writings about Schoenberg’s post-1908 music that has made precisely that claim. For more than half a century, there has been a school of thought that has stubbornly resisted the notion that Schoenberg ever really did abandon referential tonal centers.
I am marginally more sympathetic to this line of thought than I am to the approach that obdurately refuses to recognize the significance of traditional harmonic elements in his “atonal” compositions. Faced with an arpeggiated triad in a line, or a dominant seventh chord as a simultaneity, or a chromatic scale segment within a melody, advocates of this approach have not consigned these elements to oblivion.
Although I am sympathetic to parts of their argument, I find their core conclusion is – ultimately – unsupportable. Invariably, advocates of this line of thought have insisted on hearing Schoenberg’s compositions in terms of a referential tonic supported by modified versions of traditional harmonic progressions. Significant residues of past practice do persist in Schoenberg’s post-1908 works, but not these residues. To claim that referential tonics are still applicable as late as 1909 is to ignore virtually every other important aspect of the pitch language. Those who persist in assigning tonal centers to works from this period are making a complementary error to those who downplay the existence of significant residues of past harmonic practice in those same works. Both approaches miss essential aspects of the music.
I should not like to leave the impression that everyone who has looked at Schoenberg’s works from around 1908 has regarded them in the starkest of black and white terms as either “tonal” or “atonal”. There have been some writers who have understood these works as a mixture of tonal and atonal elements. For example, David Lewin, in a much admired analysis of one of the songs from Schoenberg’s Op. 15 remarked that “in general, now, I feel that tonality functions in this work mainly as one means of clarifying, enriching and qualifying a basically contextual (‘atonal’) structure.”16 And Reinhold Brinkmann in a similarly much admired analysis of Schoenberg’s Op. 11, No. 1 frequently acknowledges the importance of tonal elements in what he otherwise sees as an atonal work.17
Seeing these works as a mixture of tonal and atonal elements is certainly an improvement over seeing them as one or the other, but I still find this an unsatisfactory solution because it assumes that Schoenberg’s music is based on two different musical languages, tonal and atonal. As I propose to demonstrate, Schoenberg’s pre- and post-1908 music cannot be divided into two separate, distinct, and different musical languages. Rather, all of Schoenberg’s works from 1899–1909 are based on a single musical language. Although this language underwent significant changes, a basic core of techniques, procedures, and ideas remained in common.
In short, I believe we would do much better to dispense with simplifying slogans (“tonal” vs. “atonal”) entirely. Trying to capture the complicated reality of Schoenberg’s music with a pair of binary opposites is not just futile, it is deceptive. It impedes understanding.18
We come then, to the revisionist thesis promised in the title of this chapter (though the reader who has been following the argument to this point has probably already inferred what its essential outlines must be):
From the beginning of his career Schoenberg subjected his pitch language to a relentless process of change. Step by step, Schoenberg continually modified or transformed many of the techniques that had characterized his music at the beginning of his career. From approximately 1899 to July 1909, one must understand the pitch-language of Schoenberg’s works as comprising an ongoing extension and transformation of prior techniques, not a renunciation of them.19
In the course of the chapters to follow, we shall attempt to retrace the path of the transformation of Schoenberg’s musical language during the decade 1899 to 1909. But this is not a story of pitch language alone. As we shall see, it is also a narrative about form, motive, aesthetics, and the idea of the modern.
2 “Based on tradition”: Four Songs, Op. 2, 1899
I venture to credit myself with having written truly new music which, being based on tradition, is destined to become tradition.1
Given the continuous, restless transformation of Schoenberg’s compositional thought, any starting point we choose might appear to be at least somewhat arbitrary. Wherever we start, we are apt to feel slightly uncomfortable, as if we have entered without proper preparation into the middle of a complicated debate.
It cannot be helped: we have to begin somewhere and for the purposes of the present study there does not seem to be much point in going back to the juvenilia from the very beginning of Schoenberg’s compositional career.2 I propose instead to begin with Schoenberg’s songs from 1899 (some of which appeared in the Four Songs, Op. 2). My criterion for choosing this as the starting point is my impression that with the 1899 songs Schoenberg first began to speak in his own distinctive voice.3 To my ear, all of his pre-1899 compositions sound at least somewhat derivative, be they redolent of Wagner or Brahms or perhaps (as in the String Quartet in D, 1897) Dvořák.4 With the songs from 1899, Schoenberg’s music takes on a far more self-assured character and starts to sound – for want of a better term – Schoenbergian. At the same time, 1899 is also appropriate for a starting point because by that date Schoenberg had not yet advanced very far down the path that led to the transformation of so many of the important structural features of the pitch language he had inherited. Beginning in 1899 thus allows us to establish a clear base line for Schoenberg’s compositional practice and thought.
Rather than survey all of the songs from 1899, I propose instead to focus on just one of them and to examine it in some detail. Not that such an approach is without its dangers. Given Schoenberg’s constant striving for originality, the choice of but a single work carries with it the danger that aspects unique to that work might mistakenly be taken as normative or that important features that occur only in other works will be missed entirely. To avoid these pitfalls, I will cite similar passages from other, approximately contemporaneous works and, when necessary, will also direct our attention to important features in other songs if they are not included in the work under discussion.
|Image not available in HTML version|
Without further ado, let us turn our attention to the first seven measures of “Schenk mir deinen goldenen Kamm (Jesus bettelt)”, Op. 2, No. 2 (Ex. 2.1).5
A prominent feature of this passage is its only slightly elaborated homophonic texture. Excepting occasional passing-tone, appoggiatura, and suspension figures, Schoenberg presents a straightforward succession of chords whose surface rhythm is even sometimes (as in the piano part in the first measure) homorhythmic.6
Homophony is not the only texture that appears in Schoenberg’s early songs, but it is extremely common. Many other passages in the songs reveal themselves to be little more than elaborated homophony. Therefore, it is fair to state that in his early songs, Schoenberg customarily (though not invariably) structured his music in terms of successions of chords.7
In the tradition from which Schoenberg’s music sprang, chords were not arbitrary combinations of tones. Rather, the traditional rules of counterpoint dictate that all of the tones of a first species combination need to be consonant with one another.8 It followed that seventh chords (and other chords with dissonances) did not appear as first-species combinations. In fifteenth and sixteenth-century choral music, this distinction was reflected in the part-writing: sevenths did not function as even locally stable tones. Rather, they were treated like (and were, in fact) dissonances (passing tones, neighbor tones, or suspensions).
Long before Schoenberg began to compose, the treatment of the seventh (and other chordal dissonances) had undergone a fundamental transformation. In both theory and practice, these dissonances had evolved into something much more than tones of figuration. Chordal dissonances had become integral components of chords, even to the point of appearing within first species combinations.
Quite early on, and particularly in instrumental music, the reality of the treatment of chordal dissonances was at odds with what theory said was supposed to happen. Composers increasingly treated the seventh and other chordal dissonances in ways that were dramatically different from their origins: chordal dissonances appeared without preparation; they failed to resolve properly or at all; they were themselves subjected to prolongation and elaboration by diminution; chords with dissonances even became temporary goals, the target of resolution of even more dissonant sonorities.9
By the time Schoenberg began composing, it was entirely normal to treat chordal dissonances such as the seventh as integral components of chords, locally stable, nearly indistinguishable from the remaining tones of the chord. Schoenberg took this stage in the development in the treatment of chords with dissonances as his starting point and promptly moved forward.
Schoenberg’s chordal vocabulary differs from that of his predecessors primarily by its emphases. Like composers before him, Schoenberg’s chordal vocabulary consists almost entirely of triads and seventh chords. What is striking about his arsenal of chords is the extraordinary pervasiveness of chords with chordal dissonances. Schoenberg’s chords tend not to be
© Cambridge University Press
Table of Contents
Preface; 1. 'Atonality': a revisionist thesis; 2. 'Based on tradition': Four Songs, Op. 2, 1899; 3. The principle of incremental innovation: Verklärte Nacht, 1899; 4. Conservative song-cycle, progressive cantata: Gurrelieder, 1900-1911; 5. Programmatic music and its implications: Pelleas und Melisande Op. 5, 1902-3; 6. Consolidation: Songs, Op. 3, 1903-4; 7. Abstract form, secret program: String Quartet, Op. 7, 1904-5; 8. Referencial centres? Lieder and Fragments, Fall 1905; 9. Absolute music and its consequences: Chamber Symphony, Op. 9, 1905-6; 10. Crisis: Friede auf Erden, Op. 13, Ballades, Op. 12, and the reception of Shoenberg's music, August 1906-July 1907; 11. Motivic economy: String Quartet No. 2, Op. 10, movements one and two, March-December 1907; 12. 'Until then I lacked the strength and confidence': Two Songs, Op. 14, December 1907-March 1908; 13. Beyond triads: the first layer of Das Buch der hängenden Gärten, Op. 15, March-April 1908; 14. 'On revient toujours?' Returning to Opp. 10 and 15, June 1908-February 1909; 15. The analysis of Schoenberg's post-1908 music: Pieces for piano, Op. 11, nos. 1 and 2, February 1909; 16. 'Intoxicated by the enthusiasm': Five Orchestral Pieces, Op. 16; Piece for Piano, Op. 11, No. 3, May-August 1909; 17. The birth (and death) of new music: August 1909 and beyond; Bibliography.