Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

Transcription

1 The Evolution of Twelve-Note Music Author(s): Oliver Neighbour Source: Proceedings of the Royal Musical Association, 81st Sess. ( ), pp Published by: Taylor & Francis, Ltd. on behalf of the Royal Musical Association Stable URL: Accessed: 25/08/ :41 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org. Royal Musical Association and Taylor & Francis, Ltd. are collaborating with JSTOR to digitize, preserve and extend access to Proceedings of the Royal Musical Association.

2 7 FEBRUARY I 955 The Evolution of Twelve-note Music OLIVER NEIGHBOUR Chairman FRANK HOWES, C.B.E. (PRESIDENT) THERE HAS BEEN a tendency to treat twelve-note music as an isolated phenomenon. This is appropriate in a close study of its structural features and musical purpose, but we shall be in a better position to ask the right questions about its origin and influence if we take into account other music which shows parallel developments. Such a preliminary enquiry may affect our view of its nature and historical significance, and my subject is this first step: less the evolution of twelve-note music than its place in the general evolution of music at about the time of its emergence. Between the years I9I4 and I921 three composers seem to have arrived more or less independently at the idea that the twelve notes of the chromatic scale might yield an alternative constructive principle to tonality. These were J. M. Hauer, Arnold Schoenberg and a Russian named Golyshev. According to Herbert Eimert who wrote a theoretical book on twelvenote music in I924, Golyshev composed a string quartet in 1914 which showed dodecaphonic characteristics. These were further developed in an orchestral work of 19I9, Das eisige Lied, but unfortunately neither work seems to have been published. We must therefore confine our enquiry to the other two composers, about whom we are well informed. Hauer tells us that he discovered dodecaphonic procedure in his own music in 1919, and applied it consistently from that date. He also speaks of performing twelve-note music privately considerably earlier, but to judge from such early works as the Third Symphony of 1913, he is not using the term in any very strict sense in this instance. Schoenberg wrote his first strictly twelve-note pieces, later incorporated in his Piano Suite, op. 25, in the autumn of 192I. Schoenberg sketched a symphonic scherzo, partly based on a twelve-note theme, in the winter of I His first strictly 49

3 50 twelve-note pieces followed, in the autumn of 1921, after several years of experiment. These were later incorporated in the Piano Suite, op. 25. The root of these developments has been sought in the work of representative composers, particularly German ones, of the early years of the century. There are fairly frequent passages here foreshadowing the simplest factor of the twelve-note method, the one that gives it its name. Such evidence is not always relevant. The twelve notes are to be found in close proximity even in Wagner, for as tonality expanded its boundaries chromatic sequences arose which brought them together. There is no suggestion of a new constructive principle in this, and the most that can be said is that the mere presence of such passages prepared the way for later developments. Gradually, however, chromatic variety began to exert an attraction of its own. This is observable in some of Mahler's and Reger's melodies which reach out, without recourse to sequences, almost to the twelve-note limit before turning back, though they are harmonised within a tonal framework. Where this tendency became especially strong, as it did in Schoenberg, it affected every aspect of the music so that the framework could not hold. It is interesting that Schoenberg noted in his Harmonielehre of 9I I that he had found himself avoiding octave-doubling in his harmony' and the use of the same notes in two consecutive chords. It is one of the few comments he permitted himself that bears on the music he was writing at that time, but it is fundamental. We have seen that the experience was not confined to him, and also that the twelve-note idea, which is its logical outcome, emerged in the work of various composers. We may fairly conclude that the tendency towards total chromaticism was a natural historical development. A comparison of the methods of Hauer and Schoenberg may help to clarify its nature. Hauer found that the twelve notes could be divided into two groups of six in forty-four different ways. It has been maintained that he miscalculated, but that need not concern us. He adopted these pairs of six-note groups as the basis of his method, and called them Tropen, a term which I propose to translate as 'tropes'. Since each trope may be transposed to any pitch, any twelve-note melody must belong to one of the ' p This is made explicit only in the third edition of 1922, p

4 forty-four when divided in the centre. Further, it may be divided into two groups consisting of notes 2 to 7, and 8 to 12 together with the first note. If we keep moving one note to the end and dividing at the centre we arrive at a series of six tropes, after which we get back to the first one. Certain melodies belong to fewer than six tropes, some divisions merely giving the same tropes at different transpositions, but none can belong to more than six. Hauer uses these chains of related tropes in his compositions, but he does not necessarily preserve the identity of the melody which underlies them, for the notes of each half trope may be heard in any order. By comparison the organisation of Schoenberg's series is well known and easily described. A basic series is chosen and adhered to throughout a piece, usually without changes in the order of the notes. Its inversion, and the cancrizans of both forms, are also used. Any tran position is allowable. These rules are of little importance in themselves; their potentialities become apparent only in their application. Moreover, the application may be said to precede the rule, at any rate in Schoenberg. The rule serves to rationalise and co-ordinate; removed from its context it has little meaning. Its adoption by any composer worthy the name places it in a new context and brings out fresh characteristics. It operates in accordance with a different musical mentality and helps to develop different qualities. Here, equally, the application precedes the rule. I emphasise this point because in comparing Schoenberg with Hauer or other composers I shall be referring to him alone, not the convenient representative of a school. I hope the force of this will become evident as we proceed. It is not clear in what guise Hauer discovered the dodecaphonic idea in his own work, but his first theoretical concept was of a twelve-note melodic line without rhythm or accompaniment. This for him was melody in its purest form, the Melos, devoid of any disruptive influence such as a leading note. In composition, however, stress and rhythm could develop out of the nature of the intervals, and harmony result from melodic notes continuing to sound after they had served their melodic purpose. In its simplest form this principle produces a kind of counterpoint in which the underlying melody moves from one part to another, so that the parts themselves move one at a time. Several of the pieces in Hauer's first published twelve-note work, the Atonale Musik of 5I

5 , are of this kind. There is a passage (bars 44-59) in the waltz that concludes Schoenberg's set of piano pieces, op. 23, written at about this time, which may have been written under their influence. In a livelier version of the same technique melodic notes are taken up in other parts at different octave levels so that more movement is obtained. Chordal accompaniments to melodic lines are often similarly formed from the notes of the melody itself, much as a melodic figure consisting of an arpeggiated triad in tonal music may be harmonised by that chord itself. In all these procedures the music is derived from a melodic line and its projection in various forms. The notes of the Melos are, however, sometimes heard in a succession of chords, and a more complex kind of counterpoint is related to this treatment. The six notes of each half trope are disposed in a number of parts which pass more or less simultaneously from the sphere of one half trope to the next, so that the identity of the trope is preserved, not in linear form, but in the aggregate of several contrapuntal strands. From such parts others may be derived by the usual method of projection. The idea of the Melos is decisive for the nature of the music. Since no rhythmic character may be imposed that is not felt to arise from the inherent tension of intervals, the music tends to proceed in rather regular note values, continuity being secured by melodic correspondences. The dramatic oppositions and thematic development of sonata form are lacking. Schoenberg's twelve-note practice is too well known to need description here, and its variety is such that time would not permit it. If his fundamental rules can be more shortly stated than those of Hauer, their application is incomparably more diverse. The procedures of the two composers differ sharply, having evolved quite separately to meet different compositional needs. In Schoenberg's later expressionistic works there are passages in which certain intervals or combinations of intervals are used both melodically and harmonically as a means of tonal organisation. A few years later he developed this idea further by taking a series of notes or intervals as the sole basis of a piece of music, so that harmony and melody both became aspects of the motive, or motival in character. A method designed in the first place to provide a unifying idea from which the thematic correspondences of classical music should arise was utterly foreign to Hauer's musical aims. It was not even necessarily a twelve-note method. Some of

6 53 Schoenberg's pieces of the formative period are based on series of a different character; many years later he used similar principles in a tonal setting in parts of the Organ Variations; and in very late works serial considerations are occasionally permitted to over-rule the twelve-note idea instead of safeguarding it.3 However, when, as is usual, the method is applied to a dodecaphonic series it resembles Hauer's technique only in special cases, and then, as it were, by chance. Thus Schoenberg often divides his series into two complementary six-note groups in the manner of Hauer's tropes, though for different reasons, and when he happens to use these groups separately, one after the other, there is a momentary technical resemblance to those of Hauer's procedures which exclude doubling. But when he uses two forms of the series or two sections of the same form simultaneously in counterpoint, the order in which the notes are actually heard is not subject even to the degree of control that is exerted by the caesura in the tropes. Instead, the notes are controlled by the serial thread which, running now horizontally, now vertically, draws them all into the context of the piece, functioning as Schoenberg himself said, 'in the manner of a motive'. It is therefore the organisation, whether in Schoenberg or Hauer, that expresses the musical aims and intentions of the composer. I think the attraction of the twelve-note idea and the problem of its organisation are very well illustrated in the music of Bart6k, who was, as far as I know, the first composer to use a group of twelve notes consciously for a structural purpose. The earliest example is the third of his set of fourteen bagatelles, which were completed in I908. It consists of a melody in ternary form set against an ostinato of five different notes. The first two phrases of the melody add four more notes to these, the third five more, the fourth and culminating phrase six. At this point then, eleven different notes are used in conjunction, and the phrase leads through a crescendo to the twelfth note, which is the first of the recapitulation. There can, I think, be no question that this is a conscious piece of construction, for the idea of starting from a limited tonal field and reaching a climax in the fullest possible chromaticism became a prominent feature of Bart6k's work for the next few years. There are striking examples in Bluebeard's Castle (19I I). The idiom is sufficiently chromatic for sequences and 3 e.g. Violin Fantasy, bar 30.

7 54 scalar passages to bring the twelve notes together from time to time without special significance. But in three passages4 at the climax of the opera, when Judith is crowned and passes through the seventh door, they are clearly used purposely to produce extreme tonal tension. For instance, there is the typical Schoenbergian device of setting a melodic phrase of eight different notes against a chord consisting of the remaining four. This occurs twice, once within fifteen notes and once within fourteen. Nearly every point of climax in the Wonderful Mandarin (1919) is similarly marked, in one case by a pair of complementary chords,5 again reminiscent of the Schoenberg of a few years later. Sometimes, however, Bart6k used twelve-note groups for an exactly opposite purpose. Far from building up to his greatest tension he would begin a work in this way, as though to emphasise from the outset the tonal range of the movement which is to follow. The opening of the First Quartet (I908) presents the twelve notes within the first fourteen. In the Wonderful Mandarin an impressionistic background of eleven notes prepares for the entry of the first theme on the twelfth. The First Violin Sonata (192I) begins with a similar accompaniment, built upon ten notes, which introduces a violin melody pivoting upon the remaining two. Even so late a work as the Third Quartet (1927) has a comparable opening. The contradiction between these two uses of the idea is obvious. In the Wonderful Mandarin they occur somewhat illogically side by side. It is not surprising that in the First Violin Sonata they should have lost their identity, the twelve notes being brought together purposely fairly often, but without any apparent structural intention. In the recapitulation of the passage I have already described, the accompaniment is twice changed in such a way as to keep all the notes in play except the salient ones in the violin melody.6 Elsewhere all the notes may be found in various chordal or contrapuntal forms within as few as thirteen notes.7 So many passages of an extremely chromatic nature were not easily to be brought within a structure of tonalities. Bartok was faced with a choice. Continuing in the same direction he would have abandoned tonality and reached 4 Section 132, bars 1-3, 9- I; section 133, bars 2-3; section 136, bars 9-0o. 5 Section 82. Other examples are Section 04, bars 6-9 and Section 12 bars 1-3, in the original version. 6 Section 20, bars 1-4, 5-8, 9. 7 cf. especially Section 3, bar 6, to end of Section 4.

8 55 a position not essentially different from that of Schoenberg in HEczgewdchse (I9II), a work in which the twelve notes are almost continuously present, though it is perhaps not intentional here. A new problem of organisation would then have presented itself, as it did to Schoenberg. Instead he hesitated until he had finished his Second Violin Sonata in the following year, and then gave up the twelve-note idea. It was not only twelve-note composers whose music moved towards alternative methods of organisation to structural tonality. In the first decade of the century Debussy explored two chordal treatments. Sometimes he would write chains of parallel chords which, since they were not modified in accordance with a scale, made for vagueness of tonality. Elsewhere he would continue a single chord or chord-like figuration through several bars, the melody largely arising from it, though probably admitting some foreign notes. Such chords may have tonal roots, but they scarcely come within the bounds of structural tonal harmony. The parallel chord principle may produce a simple transposition, but they remain static and self-sufficient. Thus the piece entitled Cloches a travers lesfeuilles cannot be heard as a structure in G. It passes through a series of tonal regions which to some extent impose their own terms. These whole-tone, pentatonic or purely ad hoc combinations are allowed to melt into one another freely; only the first was ever treated by Debussy at all systematically. I shall return to it later. The idea of using a series of tonal regions of this kind, each devised in accordance with the demands of a particular moment in a composition, was developed in a very individual way by Stravinsky. His technique may conveniently be studied in the first of the three short pieces for string quartet of This is confined to one such region, a single ostinato pattern or combination of ostinatos. The cello plays only low C together with D) a ninth above it, and E7; the viola has a held D; the second violin plays a descending scale of four notes, F#, E, D# and C#; and the first violin a recurring melody on the notes G, A, B and C. The differing length of these patterns and the pauses between their repetitions cause the parts to coincide in almost every possible combination. It is as though the piece were based on a single chord which is never heard in its entirety but from which any selection may be made. This is not so where two or more notes in one part

9 56 are not doubled elsewhere, and can therefore never be heard simultaneously. In this quartet. piece the G, A and B in the first violin part can only appear successively. But with this exception we may say that the vertical coincidence of the parts, whilst in a sense more haphazard than is usual, is also more rigidly controlled within the framework of a fundamental chord. Stravinsky made wide use of this technique from the time of Petrouchka until the early nineteen-twenties; the Rite of Spring and The Wedding consist in very large part of such ostinatos. Most are based on triadic or scalar formations which suggest a tonality. Others are not, but in either case the succession of static tonal regions has little to do with structural tonality. Each ostinato assumes the function of tonality for its duration, or we may say that an ostinato is brought into being by the necessity of defining and establishing a tonal field which can only exist through the constant re-affirmation of its limits. A melody which on its first appearance seems to be extraneous to the established field must continually double back upon itself, justifying its intrusive material by repetition, so that it becomes an extension and not a negation of the existing order. Debussy achieved a more complete unity between the horizontal and the vertical in his use of the whole-tone scale, which for him was as much a chord as a scale. Owing to its lack of intervalic differentiation it offers very limited possibilities, and its single transposition brings little variety. Debussy never used it throughout a piece, even Voiles containing half a dozen bars in which the pentatonic scale is treated according to the same principles. But to composers who were moving away from structural tonality its consonant character offered an attraction, and Skryabin began to use a formation very like it in Prometheus (I909-Io) and other works of about the same date. He arrived at this by piling up fourths of various kinds to form a so-called synthetic chord, but we have here another example of a tonal field which is equally chord and scale, even though its application often shows its chordal origin very clearly. The Prometheus chord comprises a whole-tone scale in which one note has been raised a semitone. This single irregularity makes a series of transpositions available, and correspondingly increased musical resources. This chord was soon largely abandoned in favour of other six-note combina-

10 57 tions with similar properties, but with two notes differing from the whole-tone scale instead of one, and therefore offering more scope in composition. Different sections of a work are sometimes based on different chords. These do not necessarily appear complete at every point in the composition; any selection may be made and disposed in any harmonic or melodic order. Different transpositions may overlap and lead to new combinations. Extraneous notes, sometimes completely absent, are sometimes frequent, though certain preferences are observable. Where the usage is strict, the flexibility and mobility of the technique are reminiscent of Schoenberg. One of Skryabin's favourite devices is particularly so: his use of a group of notes as a pivot between two transpositions of his chord. His Seventh Sonata, for instance, is based on the commonest of his scales: C, Db, E, F#, A, Bb. The main subject of the work makes great use of the triad with both major and minor thirds, a chord that is not generally emphasised in his music and helps to give this sonata its marked individuality. It can be derived in two ways from the scale I have quoted, so that one such chord can be interpreted as belonging to two transpositions, and can be heard with two different pairs of notes to make up the complete forms.8 There is a striking parallel here with Schoenberg's habit of punning on a group of notes common to two forms of his basic series. It is plain that all these experiments in the direction of a new kind of tonal field equating the horizontal and the vertical run the risk of static and monotonous results. The more complete the equation, the greater the risk. This is so much so in the whole-tone scale as to render it of very restricted practical use. In his last works Debussy turned away from this and similar combinations, their influence surviving only in the occasional harmonic use of melodic motives.9 Skryabin fares better with 8 Bars I2-14 provide a good example. Four related forms of the fundamental chord are used, based on C, A, F: and D:. Bar 12 is on A, except for the duration of the last two of the triplet semiquavers, where it is on C, the F# chord in the right hand acting as a pivot. The first half of bar 13 is on A, the second half on F#, the left hand punning with the previous bar. The first half of bar 14 is on A, the second half on D$, except the last two semiquavers which return to the C form. This passage is, I think, misunderstood by Zofia Lissa in her admirable article 'O Harmonice Aleksandia Skrjabina' in Kwartalnik muyczny (Warsaw), zeszyt , p e.g. Etude no. 9, bars I and I.

11 58 his more varied scales, but finds certain licences imperative. The very late works of are already considerably freer in general than those of 91 o-i. Stravinsky, whose strictest ostinatos do not perfectly equate the dimensions, was able to draw upon his method for much longer, and its traces remain in his later work. The apparent paradox disappears when we recall that each composer benefitted artistically from these techniques and required nothing more of them than they could provide. Their appearance at about the same time in the work of a number of composers indicates something more than a coincidence; a historical movement comparable to that towards complete chromaticism. Each of these two tendencies contains a potential remedy to the defect of the other. Chromaticism does not suffer from severely limited tonal possibilities; the organised field imposes the limitations that are essential to all music. For this reason they become permanent factors only in the music of Hauer and Schoenberg, where they meet. Hauer does not, perhaps, escape every difficulty. The whole-tone scale and Skryabin's chords are analogous to half tropes in that all are six-note groups, and Hauer's practice often emphasises this. Theoretically the tropes should allow great tonal variety, but when he uses a succession of half tropes as quasi-chordal entities he tends to chose forms whose consonant character makes for music no less static than that of Skryabin, who shows remarkable technical resource within a single six-note region. In his contrapuntal practice, on the other hand, Hauer often weakens the twlve-note idea by constructing additional parts from notes already heard in another octave. It is in Schoenberg that the tendencies find a more complete equilibrium. The fixed order of the notes in the serial method might seem a long way from the techniques we have been considering, the six-note groups resembling superficially the freely handled chromatic scale as it appears in Schoenberg's expressionistic works, rather than the twelve-note series. The chromatic scale, however, is a special case since it represents the complete tonal field of western music and lacks any limiting factor unless some order is imposed upon it. It is one of the functions of the series to do this; Schoenberg was sometimes prepared to relax the order of the notes within a section of a series because the field was then sufficiently characterised by its own limits. We have seen that the series

12 59 unifies the horizontal and the vertical. This is achieved in such a way that there are analogies not only with the six-note groups, but also with Stravinsky's methods. Schoenberg's serialism was largely motival in origin and purpose, so that the musical idea is in a sense equivalent to the field in which it moves. The same is true of Stravinsky's ostinatos, and if the means employed by the two composers are very different, they at least share the principle of constant repetition. From the technical point of view, Schoenberg emerges from this enquiry as the central figure of his time. Many of the most personal and original elements in the music of Hauer, Bartok, Stravinsky and Skryabin are paralleled in his work and nowhere else. I have suggested that the reason why he did not find modification in the direction of tradition necessary, as they did, was that the two main tendencies that he combined proved complementary to one another. An additional reason follows from the nature of the serial solution. One need go no further than the vertical aspect of the music. The other methods of organisation impose very severe restrictions here, whereas the extreme flexibility of the series, and the frequent simultaneous use of more than one form, make possible any vertical combination whatever. In his twelve-note works, as in those that preceded them, Schoenberg had no guide in his harmonic procedure except his innate musical sense, in other words, his experience of traditional practices and his feeling for their potential development. In this sense his music is entirely unsystematised and relies heavily on the past. It is at once more and less revolutionary than that of his contemporaries. It could not have returned to a tradition of which it was already so completely a part. Many people believe that some form of dodecaphony will become the universal musical technique, replacing tonality. Schoenberg's fusion of historical trends might seem to support such a view, but I do not think that speculation of this kind is very useful. I should prefer to put forward another interpretation of Schoenberg's position. In many ways it resembles that of Bach two centuries earlier. Both composers were profoundly traditional in their artistic aims. From the strictly technical, though not the aesthetic point of view, both may be said to include their contemporaries. Both look far forward too, exerting influence, though not immediately, in proportion to the wide implications of their musical methods, whilst remain-

13 6o ing essentially inimitable by the same token. Thus Schoenberg's influence, as far as it can be assessed at present, has been analytical. A very high proportion of representative composers now writing have learned from him, but they have not received a complete system. They have taken according to their differing needs and therefore stand in varying relation to him, rather as his contemporaries did, though the fact of influence places them nearer. They have not, of course, re-discovered the historical factors that gave rise to dodecaphony, but have selected elements from Schoenberg's total concept. The situation cannot be understood without reference to musical aims. I have emphasised the fact that Schoenberg thought of his series as motival in character. His method evolved in accordance with his genius for close musical thinking in the tradition of the Viennese classics and Brahms. Such composers are uncommon and it is not surprising that none of his followers has been so highly gifted in this way. Some have adapted his method to more purely lyrical ends, others to a new kind of strict counterpoint; all have used it in a more restricted manner. More recently the series has been subjected to elaborate permutation as an end in itself by composers whose pronouncements show that Schoenberg's musical intentions are beyond their comprehension. There is no reason why such a misunderstanding should not prove fruitful, but the technique assumes an entirely different function here. Such developments take us beyond our terms of reference. It remains to be seen whether they will eventually modify our view of the procedures with which we have been concerned this evening. (Mr. Ian Kemp played as illustrations on the pianoforte the second piece from Hauer's 'Atonale Musik', the fifth of his 'Klavierstiicke'. op. 25, and the third of Bartok's 4 Bagatelles.) DISCUSSION The Chairman welcomed the paper because of its lucid treatment of a topic away from the track of musical research. The Lecturer had related his subject to evolution which provided the best explanation of all modern techniques. But this might imply a philosophy of musical history in which no composer was ever himself, but always the predecessor or successor of another. Whenever the Lecturer referred to

14 I doctines such as that of the limitation of flux, this made him (the Chairman) feel hopeful, but when composers were shown as becoming systematic, or beginning to think tidily, he felt uncomfortable. He wondered whether the Lecturer might not be 'having it both ways' when he said that the twelve-note system brought both freedom and necessary limitation. The principle of repetition was surely inherent in all music, but an arbitrary twelve-note series tied the composer to an eternal ostinato, which seemed either the beginning of a very new development or of bankruptcy. The Lecturer replied. that he had been less concerned with a defence of the serial principle-though it could be defended -than the fact that it included most of the important developments of the period c. I This suggested such a fusion to be a natural development. Mr. Matyas Seiber said that the emergence of a twelve-note technique was an historical necessity, not just an intellectual pastime. It had saved chromaticism from complete chaos. No use of ostinato was to be found in Schoenberg. The twelve notes were a reservoir from which anything might be drawn. It was also true that the harmonic dimension was not subject to rigid control. Webern had tried to systematise this aspect of the music. He realised that all this might sound highly theoretical but in fact was not more so than composing in C major.