CHAMBER MUSIC Selections from Essays in Musical Analysis DONALD FRANCIS TOVEY DOVER PUBLICATIONS, INC. Mineola, New York Bibliographical Note This Dover edition, first published in 2015, is a republication of Chamber Music: Essays in Musical Analysis originally published by Oxford University Press, New York, in 1989. Library of Congress Cataloging-in-Publication Data Tovey, Donald Francis, 1875–1940. Chamber music : selections from essays in musical analysis / Donald Francis Tovey. p. cm. “This Dover edition, first published in 2015, is a republication of Chamber Music: Essays in Musical Analysis, originally published by Oxford University Press, New York, in 1989.” eISBN-13: 978-0-486-80141-4 1. Chamber music—Analysis, appreciation. I. Title. MT140.T69 2015 785.009—dc23 2014027844 Manufactured in the United States by Courier Corporation 78449501 2015 www.doverpublications.com CONTENTS 1. Chamber Music, a general survey (1928) BACH 2. The Clavierübung (1936) 3. Aria with Thirty variations (The ‘Goldberg’ Variations) (1900) 4. A Listener’s Guide to Die Kunst der Fuge (1936) 5. Toccata in F sharp minor (1919) HAYDN 6. Pianoforte Sonata in E flat, No. 1 (1900) MOZART 7. Quintet in E flat (K. 452) for pianoforte, oboe, clarinet, horn, and bassoon (1900) 8. Adagio in E major for violin (K. 261) (1900) BEETHOVEN 9. Thirty-three variations on a waltz by Diabelli, for pianoforte, op. 120 (1900) 10. Sonata in A major, for violin and pianoforte (dedicated to Kreutzer), op. 47 (1900) SCHUBERT 11. Blumenballade: ‘Viola’ (1900) SCHUMANN 12. Novelette in F sharp minor, op. 21, no. 8, for pianoforte (1900) 13. Quintet in E flat, op. 44, for pianoforte, two violins, viola, and violoncello (1900) CHOPIN 14. Observations on Chopin’s Etudes (C sharp minor, op. 25, no. 7, D flat major (for the Methode des Methods), No. 3, F minor (ditto), No. 1, and C minor, op. 25, no. 12), for pianoforte (1900) 15. Impromptu in F sharp major, op. 36, for pianoforte (1900) BRAHMS 16. Variations and Fugue on a theme by Handel, op. 24, for pianoforte (1922) 17. Variations on a theme by Paganini, op. 35, for pianoforte (1900) 18. Quartet in G minor, op. 25, for pianoforte, violin, viola, and violoncello (1901) 19. Quartet in A major, op. 26, for pianoforte, violin, viola, and violoncello (1901) 20. Quartet in C minor, op. 60, for pianoforte, violin, viola, and violoncello (1901) INDEX CHAMBER MUSIC 1. A GENERAL SURVEY (1928) LIKE most classical terms, chamber music has meant different things at different periods, and has developed by evolution, it is a mistake to try to reduce early and late products of an evolutionary process to a common definition. There are theoretical objections to any limit that can be proposed. These may be removed by a statement which clearly shows the relation of what is excluded from the main scope of this essay to what is included. The great change which came over the whole art of music after the middle of the eighteenth century affected chamber music no less profoundly than it affected opera. Nobody would quarrel with a history of opera for beginning with Gluck, so long as it did not wholly ignore his antecedents, archaic, prophetic, and decadent. But the kind of operatic art from which Gluck revolted rested on principles so radically and, to our notions, so obviously wrong, that all attempts to revive it must savour of antiquarianism. This is not the case with the chamber music of the earlier eighteenth century ; its principles, though now obsolete, were consistent and true to the nature of the instruments, and its masterpieces can never become antiquated. And their quantity is enormous. The whole mass of the chamber music of Bach, Handel, the Italian violin masters, and the French clavecinists must be far more voluminous than the sum-total of important chamber music from Haydn onwards, even if vocal music be excluded. And to review it, or even to read a review of it, is a task at which librarians might quail. Fortunately for the reader, though it cannot be ruled out on artistic grounds, it is all based on two principles which are radically opposed to those of the later classical chamber music. Accordingly, my first part is devoted to illustrating the theory and practice of the continuo in the earlier chamber music, with occasional remarks on the other archaic principle, the mechanical use of 4-foot and 16-foot stops on the harpsichord and organ, the instruments to which the continuo was entrusted. Though much has been written about the continuo, no writer has hitherto shown its relation to the aesthetics of later chamber music. But unless this relation is clearly understood, we cannot properly understand the revolution effected by Haydn and Mozart, nor fully appreciate the qualities of purity and euphony in chamber-music style. The continuo and the use of 4-foot and 16-foot doublings represent important musical instincts. To ignore or misunderstand them is to deprive the earlier chamber music of its euphony. In the later chamber music the continuo instinct sometimes reasserts itself (often without the composer’s realizing its origin) as an impurity of style; especially when he is in the habit of composing at the piano. T C P HE ONTINUO ERIOD For general purposes, chamber music may be defined as instrumental music written for a group of individual performers, and intended to be heard for its own sake in such rooms as are to be found in private houses. Dance music, and music intended ‘to accompany the clatter of dishes at a princely table’, exclude themselves from the category of music intended to be heard for its own sake. The size of the room is not a matter for rigid definition ; the hundred-and- fifty years during which the classics of chamber music were composed were a period of royal and aristocratic patronage, and the rooms for which the music was designed were the rooms of palaces. And it is not an unmixed evil that chamber music should be heard in halls that are too large for it. The necessity can arise only because of a remarkable public demand for the highest and most spiritual form of music, and the acoustic disadvantages have a distinct value as a stimulus to the imagination. The listener naïve enough to expect the ff of Schubert’s D minor Quartet to sound loud in a concert-room holding an audience of 2,000, learns in five minutes to prefer spiritual to material values in music, if he can learn anything. Nevertheless, the classical idea of chamber music implies bigness as well as intimacy, and the listener is not enjoying the normal effect of a trio or quartet unless the sound is filling the room. This classical notion of bigness determines the art-forms ; no classical chamber music is merely lyric. Beethoven’s Fugue in D major, op. 137, for string quintet takes less than four minutes to play, and begins and ends softly; but fugues are not lyrics, and Beethoven’s sketches for this opusculum are entangled with ideas of a fugal opening which afterwards took shape in the scherzo of the Ninth Symphony. Other short pieces of chamber music are either large sets of variations or fragments of projected complete sonata-schemes. Schumann’s Märchenerzählungen and Dvořák’s ‘Dumky’ Trio are exceptional groups of lyrics by composers who otherwise accept the classical view that a chamber work must contain at least one movement in developed conata form. This rule is a natural result of the feeling that when two or more people are gathered together to play music, they may as well take the opportunity of doing more than can be done by one person. In Germany it is even considered a solecism to call a duet concert a chamber concert ; chamber music is held to begin with trios. An upper limit has not been assigned. Mozart’s Serenade in B flat for thirteen wind instruments is, for reasons which will appear in the course of this essay, within the limits of classical chamber music, though even the most experienced wind players will feel in its performance that a conductor adds much to the comfort of a group that is twice the size of the wind band of any symphony before Beethoven. Mozart’s group of thirteen, however, does not include trumpets and drums, that is to say, it includes nothing which is either enormously stronger in tone than the rest of the ensemble or enormously inferior to it in musical resource. The trumpet in Saint-Saëns’s little Septet is almost the only instance of the introduction of such powerful orchestral artillery into chamber music in a classical style, and here the classical style is jocosely archaic. The double-bass (or, as in Mozart’s Serenade, the contrabassoon) does undoubtedly bring into chamber music the question of inferiority in musical resource. This will be discussed later, but here we may conveniently note that inferiority in musical resource is not to be measured by the amount of conspicuous display, whether in melody or flourishes, but by the necessity for the use of the instrument. And the double-bass may obviously be very necessary for the support of so large a group as eight or more instruments. The criteria thus far indicated for classical chamber music seem simple and not formidably exacting. Every part is to be necessary, and the ensemble is to be complete in itself. Therefore an instrumental group must be capable of making a coherent ensemble, so that, whatever the art employed in combining sounds on different planes of tone, the chamber music style does not encourage the use either of an instrument which cannot be allowed to use its normal strength, or of one which cannot make itself heard without constant strain on its own part and constant repression of the other instruments. Obvious as this all seems, there is a whole classical period in which none of these criteria can be taken literally. The main life-work of Haydn may be regarded as devoted to organizing the art of making groups of instruments cohere as effectively as groups of voices. Instruments do not normally so cohere, and the commonest form of bad instrumentation is that of the composer whose orchestral and quartet writing is choral and organ music distributed among combinations of instruments. The worst feature of such a style is that it does not sound obviously bad, except to a highly experienced and discriminating ear, which resents the feeling of wasted opportunity as immediately as an insulting cacophony, for the grammar of the chorus and the organ has a universal validity in virtue of which its use as a basis for instrumentation may best be vindicated and condemned by the term ‘fool-proof’. But during the period of supreme mastery of the chorus and the organ which culminated in Bach and Handel, nothing is more significant than the absolute and instinctive renunciation of all attempts to make groups of instruments cohere in the same way as groups of voices. Haydn showed how to make them do so as effectively, but on principles fundamentally different; this he could never have done if the masters before him had not cultivated their sense of the different planes of instrumental sounds with minds untroubled by the problem of focusing the planes together. Nor were they wrong in neglecting this problem. Difference of plane gives rise to an enormous number of the highest aesthetic qualities in all arts ; and there is no imperfection in a scheme of instrumental music in which the main parts are left completely free to execute polyphonic designs, while the task of supporting these designs with a coherent mass of harmony is relegated to a continuo player extemporizing on a suitable keyboard instrument from a figured bass. In modern performances of such nnisic grave errors may arise from ignorance of two matters which were common knowledge to all Bach’s musical contemporaries : first, the art of filling out a figured continuo part, and, second, the general characteristics of early eighteenth-century musical forms. Without a careful explanation of these matters it is impossible to appreciate the immensity of the task achieved by Haydn and Mozart, an immensity hardly understood by any of their contemporaries, in obtaining the harmonic background from the principal instruments themselves without the aid of any permanently subordinate part. Until this was achieved, chamber music could not be defined as an ensemble to which each player contributes an individual part. For not only did the individual parts themselves not claim to be complete without the impersonal continuo, but the very notion of an individual part was purely musical, and independent of the number of performers. Bach’s organ trios are trios for two manuals and a pedal-board, to be played by one organist. But his trio in Das Musikalische Opfer, for flute, violin, and continuo, realizes its intended effect only when four instruments play it, viz. the flute, the violin, a ’cello to play the bass of the continuo, and a keyboard instrument to fill out the harmonies according to the figures. Without the ’cello to lift the bass of the continuo on to its proper plane as one of the three trio parts, the mere filling out on the keyed instruments fails to complete a trio in Bach’s sense. In modem times four great errors arise in the treatment of the continuo. First, there is the total omission to fill it out—an error which so little impairs Bach’s gorgeous polyphony that most listeners devoutly admire the resulting coldness of tone as a mysterious classical virtue. Secondly, there is the filling-out with a tone on the same plane as the real parts—an error impossible to a harpsichord player, who had but to draw the right stops for an accompaniment instead of playing on solo or tutti registers. The third error is the attempt to work out the figuring in a kind of polyphony that competes with the real parts. The fourth and worst error is the filling out with parts that avoid doubling or otherwise colliding with the main lines. This error is assiduously inculcated by theorists who also fiercely denounce the substitution of the piano for the harpsichord. But the piano is, as Philipp Ernanuel Bach pointed out, even better than the harpsichord for continuo work, only it must be remembered that the contemporaries of Bach’s sons knew both instruments, and could make no mistakes as to the kind of piano touch required. A piano filling-out can be made into an ideally soft background, but the instincts of the average modern pianist do not incline that way, and nothing can be more banal than the intrusion of an ordinary pianistic mezzo-forte into the background tone of an eighteenth-century polyphonic design. The error of a too elaborate polyphony in continuo work rests upon records of Bach’s own marvellous freedom therein, but a little practical experience shows that ordinary skill in handling harmonic progressions will, with the aid of an occasional imitation of four notes of a scale in thirds or in contrary motion, cause the naïve listener to gasp with astonishment, and spread abroad the legend that everything has been decorated in six-part canonic counterpoint. As a fact in practice and aesthetics, nothing is more miserable than the attempt to fill out Bach’s polyphony with additional individual counterpoints. The efforts of Robert Franz were thought, in their own day in the middle of the nineteenth century, to be indistinguishable from Bach’s own real parts. At the present day it is hard to understand how anybody could have thought such bundles of dropped stitches and loose ends acceptable on any theory, though worse things have been perpetrated in more recent editions of Bach. The fourth error is all the worse for the erudition that is still spent on its cultivation. Franz, in his edition (Breitkopf and Härtel’s Kammermusik- Bibliothek) of Bach’s trio (from Das Musikalische Opfer), condemns the continuo realization by Bach’s masterly pupil Kirnberger as a setting that ‘tramples on all the fine flower of Bach’s polyphony’. Unfortunately it is Franz’s own setting that tramples on the fine flower, systematically, and exactly as he differs from Kirnberger. Franz thinks of the filling-out as on the same plane of tone as the main parts, or, to conceive the matter conversely, he thinks of the ensemble as it would sound if the flute and violin were played on one piano and the continuo on another. In such conditions a collision between a plain chord and the same chord plus a long appoggiatura would be hideous, as any one may