The Origins of Beethoven's D Minor Sonata Op. 31 No. 2Author(s): Barry CooperSource: Music & Letters, Vol. 62, No. 3/4 (Jul. - Oct., 1981), pp. 261-280Published by: Oxford University PressStable URL: http://www.jstor.org/stable/736619Accessed: 25/07/2010 23:40

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay 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 athttp://www.jstor.org/action/showPublisher?publisherCode=oup.

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

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to Music &Letters.

http://www.jstor.org

THE ORIGINS OF BEETHOVEN'S D MINOR SONATA Op. 31 No. 2

BY BARRY COOPER

No autograph score survives for Beethoven's only D minor sonata, Op. 31 No. 2, and there is very little sketch material either. Indeed Hans Schmidt's comprehensive index of Beethoven's sketches' contains no references at all to sketches for the sonata. The one sketch hitherto recognized, a plan for the first movement, found on f. 65v of the 'Kessler' Sketchbook, was transcribed in full by Gustav Nottebohm,2 since when it has been the subject of scrutiny by several writers,3 but statements by Nottebohm and more recently by Richard Kramer4 that no further sketches exist for the sonata sound authoritative and conclusive. Nevertheless the 'Kessler' Sketchbook actually contains a series of independent sketches that show such coherence with each other and with the final version of the piece that they are indisputably part of the sketch process for this sonata, albeit at a very early stage; there are also some sketches, originally intended for other pieces, which can be related to the working-out of the sonata. Chief amongst these sketches are the following:

(i) f. 3r, staves 5-16: ideas for three movements, two in A minor and one in A major, probably intended for a sonata in A minor;

(ii) f. 66r/3-14: a series of concept sketches for keyboard movements in B flat major, D major and D minor;

(iii) f. 68r/3-4: an eight-bar fragment of a keyboard piece in E flat that contains figuration resembling the finale of Op. 31 No. 2;--

(iv) f. 89'/9-16: some ideas perhaps intended for a sonata in C; (v) f. 90"/3-12: a draft of a D minor sonata movement, followed

on staves 13-16 by a D minor passage in 6/8;

l 'Verzeichnis der Skizzen Beethovens' (hereafter 'SV'), Beethoven-Jahrbuch, vi (1965-8), 7-128.

2 Ein Skizzenbuch von Beethoven, Leipzig, 1865 (trans. Jonathan Katz in Two Beethoven Sketchbooks, London, 1979), pp. 27-28. The 'Kessler' Sketchbook (SV 263), preserved in Vienna, Gesellschaft der Musikfreunde, A 34, is published in facsimile, with introduction and index by Sieghard Brandenburg, in Publikationen der Sammlungen der Gesellschaft der Musikfreunde in Wien, ii, Munich & Salzburg, 1976.

' See especially Rudolph Reti, Thematic Patterns in Sonatas of Beethoven, New York 1967, pp. 201-4, and Oswald Jonas, 'Beethoven in der Interpretation', Der Dreiklang, vi (1937).

4 The Sketches for Beethoven's Violin Sonatas, Opus 30 (unpublished dissertation), Princeton University, 1974, p. 401. This dissertation contains the best and most detailed study of the 'Kessler' Sketchbook to date.

'Transcribed Nottebohm, op. cit., p. 27.

261

(vi) f. 91v/5-6: a four-bar fragment in 3/4 beginning in C major; (vii) f. 95r/15-95v/13: an extended sketch for the exposition of a

movement in A minor labelled 'Sonata 2da';' (viii) f. 96v/1-4: a deleted sketch intended for part of the C major

slow movement of Op. 31 No. 1. Taken as a group along with the final version of Op. 31 No. 2,

these sketches reveal several things about the sonata that have previously passed unnoticed. The one well-known sketch for the sonata-the one on f. 65' transcribed by Nottebohm-itself contains several unexplained features. Why is it headed 'Sonata 2da' when no sketches for the first sonata of Op. 31 were to appear until over 50 pages later on f1. 91'? Nottebohm's two theories- either that the sonata was to be paired with the already completed Op. 28 sonata in D, or that the words 'Sonata 2da' were added to the sketch some time later-are unconvincing and do not explain the other puzzling features. Why did Beethoven suddenly sketch the movement in the middle of work on Op. 30? Why did he not make any further sketches for the sonata during the two or three months in which the sketchbook probably remained in use? Why did he make another extended sketch for a D minor movement on f. 90' before returning to his initial ideas? Why, after work had begun on Op. 31 No. 1, did he make three further sketches-two in E flat major and one in A minor (ff. 93', 95', 95')-each designated as a sonata No. 2, when he already has a much more promising second sonata outlined on f. 65'?

There is a simple explanation that can account for all these features-the sketch on f. 65" must have been entered only after the rest of the book had been filled. Thus, while working on the C minor violin sonata, Op. 30 No. 2, Beethoven filled the top two staves of ff. 65' and 66' with unrelated material (a piano sketch in D on f. 65' and some ideas for a replacement finale for Op. 30 No. 1 on f. 66r). At this point he had already finished sketching the first three movements of Op. 30 No. 2 and was about to begin detailed work on the finale, for which purpose he preferred to have a completely fresh double-page spread on which to work. Thus he turned over to f. 66`, the rest of ff. 65' and 66' remaining blank until the sketchbook had been completed. Only then did he fill up the two pages with the sketch for the D minor sonata on the left and some other material which, as will be seen, is also related to the sonata, on the right. Such back-tracking to empty spaces is entirely consistent with what is known of Beethoven's normal sketching habits, and it is well known that he tended to work on double-page spreads rather than single leaves or sides. It was also not unusual at this date for him to leave a whole page-opening largely blank: a zood examDle occurs in the 'Kessler' Sketclibook on ff. 42'-43'

6 Ibid., pp. 35-36.

262

(immediately before the start of detailed work on Op. 30 No. 1), where only the first three staves of the left-hand page and the first four of the right contain any sketches. And if this page-opening is still not quite as empty as the one on ff. 65\-66r would have been, this merely explains why Beethoven chose the latter for making his sketch for the D minor sonata.

The next sketch to be considered is the one for a D minor sonata movement on f. 90' (Ex. 1 ) .7 Nottebohm passes over it without comment, while Kramer suggests that staves 3-12 may have been intended for Op. 31 No. 2, and Brandenburg's index also links it with Op. 31." But no author has pointed out the close connections between this sketch and the first movement of Op. 31 No. 2, although the relationship is as close as, for example, the one in the case of the 'Eroica' Symphony and its initial sketch plan on pp. 44-45 of the 'Wielhorsky' Sketchbook (SV 343).

Ex.I [staves 3-41

rAs .~~

15-61

r . r r r

M.G.?] [7-81

erster Theil in B

2da parte colla repetizione [-] e la prinma parta [9-10]

[K _ _ _ __E 1 -- e - a -- e~~~~~~~~~~

7 Angle brackets indicate deleted material. ' Kramer, op. cit., p. 45; Brandenburg, op. cit., p. viii.

263

_63 _ / _ i I rI I I

Like the sketch on f. 65', the one on f. 90 is of a type encountered occasionally in the very early stages of Beethoven's sketching of a movement. Recent literature has not yet produced a standard term for this type of sketch, but phrases such as 'large-scale plan', 'telescoped draft' and 'tonal overview', all of which have been used, are suitable enough; in such sketches Beethoven drafts the main corner-stones of the movement- principal themes, key centres and so on-but omits the transitional material. The sketch here begins with a 'Mannheim rocket' motif accompanied by a repeated dominant internal pedal; it continues

264

with a descending quaver pattern leading on to a dominant chord, and an answering phrase completes the theme. What Beethoven intended next is uncertain, but a possible continuation would be the opening figure transferred to the left hand. This possibility must certainly have occurred to Beethoven at some stage, for in this form the passage would be very similar to bars 21-22 of the final version of Op. 31 No. 2: in these bars the six-note arpeggio motif has simply been reduced to its four notes, and the repeated dominant replaced by an oscillating triplet figure:

Ex.2

The next part of the movement to be sketched was the second subject. Beethoven first decided the key, with the words 'erster Theil in B' indicating that the exposition was to end in B flat; then he squeezed in a theme in that key, labelled 'M.G.' (Mittel Gedanke) or possibly 'M.T.' (Mittel Theil)-a term found elsewhere in the sketchbook,9 always to denote the second subject. The idea of taking the exposition to B flat is interesting in that this would have been the earliest sonata exposition (apart from Op. 31 No. 1, which was still unsketched) in which Beethoven closed in a key other than dominant or relative major. In the final version of this sonata the second subject is in A minor, but the use of B flat major was transferred to the slow movement. This is not the only sonata where Beethoven was undecided at first whether to employ a particular key for the second subject or for the slow movement: in the 'Waldstein' Sonata there is an early sketch that shows the second movement in E major, the key eventually used for the second subject of the first movement; and amongst the early sketches for Op. 31 No. 1 in the 'Kessler' Sketchbook (f. 92r/3-4) is a concept sketch in B major (the key of the start of the second subject in the final version), which clearly seems, from its style and its position on the page, to be intended as an idea for the slow movement of the sonata. Thus in Op. 31 No. 1 and the 'Waldstein' Sonata a key-centre was transferred from slow movement to second subject, whereas in Op. 31 No. 2 the position is reversed. As for the idea of moving from D minor to B flat major within a sonata-form movement, it was resurrected some twenty years later in Beeth- oven's next big D minor sonata-form movement-the first move-

9 See f. 3Y/15 and f. 747/1.

265

T^he exposition in the sketch on f. 90' ends with repeated B flat chords, which are developed after the double bar into A minor, another key whose position was to be shifted in the final version, this time from development section to second subject. These repeated-chord figures anticipate, and indeed help to explain, a similar idea in the recapitulation of the final version (bars 159-68). In this final version the chords seem to have little relevance to the rest of the movement (unless it be to the repeated minims at the end of the exposition), but they can now be seen as a borrowing from this sketch, where, as in the final version, they lead into rapid arpeggios. An interesting feature of the left-hand accompaniment to the arpeggios in the sketch is its anticipation of a similar motif in the overture to Egmont.

The beginning of the development section is marked '2da parte colla repetizione e la prima parta'. The term '2da parte' refers to the development and recapitulation together (without the coda), and the way the whole phrase is laid out suggests that Beethoven decided straightaway that the second part should be repeated-a relatively unusual occurrence in his sonata-form movements and one not carried out in Op. 31 No. 2 and then almost as an afterthought made a note that the first part should also be repeated as usual. But the fact that he felt it necessary to add the words 'e la prima parta' shows that he must have considered the possibility of repeating just the second part; and this possibility, however unconventional, was indeed put into effect in his next minor-key sonata, the 'Appassionata', Op. 57, in the finale of which only the second part is repeated.

The development section in the sketch leads to a half close in D minor. Then comes a remarkably bold stroke, at the point in the movement where Beethoven was customarily most daring- immediately before the recapitulation. A slow section in 3/4 time and marked 'dolce', beginning in D major, is inserted before the resumption of the allegro. The new theme is based on the motif immediately preceding it, but it is unrelated to the rest of the movement; it acts as a much-needed breathing space between the incessant forward motion of the previous music and the return of the allegro tempo (marked 'dopo l'allegro di nove'). This leads to the end of stave 12 on the page, and the remaining four staves are taken up with a sketch in 6/8 time. Kramer regards this as independent of the preceding sketch,"' but it is presumably intended for the coda of the movement since it is a transformation of the opening theme-a technique Beethoven employed in certain other codas (see, for example, the finale of the Third Piano Concerto). This transformed theme begins a 'closed' binary form that leans from D minor to F major in the first part and then back

10 Op. cit., p. 24.

266

to D minor in the second; the presence of double bars implies that both parts would be repeated, although repeats are not marked. In the final version there is no such elaborate coda (the coda is instead one of Beethoven's shortest); but the idea of a closed binary form within a coda was later resurrected-once again in the finale of Op. 57. The theme itself, meanwhile, is close to that of the bagatelle Op. 119 No. 5, the only sketch for which appears earlier in the 'Kessler' Sketchbook on f. 59.11

Taken as a whole, the sketch on f. 90' could be realized into a most impressive movement, and with comparatively little effort since so much detail is already present. It contains many novel features, which clearly point the way towards Beethoven's 'second period' and which in several cases were actually taken up in later works such as the 'Appassionata' Sonata, Egmont and the Ninth Symphony. Yet the movement contains inherent weaknesses that could not easily be removed through Beethoven's ordinary sketching methods: the opening 'Mannheim rocket' is far too conventional, looking backwards to Op. 2 No. 1 and beyond, rather than forwards; the second subject is rather lightweight and would have difficulty sustaining melodic flow and rhythmic drive in the kind of movement Beethoven obviously had in mind; and the slow interruption in D major had to be made more relevant, somehow, to the rest of the movement. The obvious solutions to these problems only throw up further difficulties: if the slow section in the major is prepared by adding a similar slow passage at the beginning the key structure is upset-a fast movement in D minor cannot begin with a slow, lyrical introduction in D major,'2 and anyway the slow passage would still not relate to the fast sections; having a slow introduction would also destroy the element of surprise-the unexpected lyricism-that Beethoven desired at the start of the recapitulation; and rejecting the 'Mannheim rocket' opening would remove the whole starting-point of the movement he was trying to create.

Beethoven was clearly immediately aware of the difficulties, for he did not develop the movement through any further sketches on the remaining six leaves of the sketchbook. Instead he began detailed work on a different sonata-Op. 31 No. 1; and his ideas for continuing what was to be the Op. 31 set with a second sonata in E flat major or A minor show that he had at that stage abandoned the D minor one, which could make procress only after a comDlete

" Beethoven probably did not often consult his early sketchbooks in his later years, but he may have looked through the 'Kessler' Sketchbook around 1821: as well as this bagatelle, the main theme of the first movement of Op. 111 appears in it, as does a theme resembling that of the second movement of the same sonata. Both of the latter are quoted in Nottebohm, op. cit., pp. 19 & 36.

" The 'Kreutzer' Sonata, Op. 47, is the exception that proves the rule: the slow introduction stays in the major for only four bars before moving to the tonic minor, and it does not return at the end of the development section.

267

reconsideration of its elements. Thus when he did return to it, after completing the sketchbook and having to resort to the blank space on f. 65v, it must have been after giving considerable thought to the matter. The end result, though familiar enough today, is in fact a model of how to solve several conflicting compositional problems without compromising the essence of the movement, and the main solutions are already present in the sketch on f. 65v, after which it would be simply a matter of sketching in the details. Hitherto, commentators have tended to regard this sketch as a sudden inspiration-a kind of written-down improvization that formed Beethoven's very first thought on the sonata;`s but it can now be seen that, however primitive the sketch may seem when compared with the final version, it represents a major advance on the previous sketch on f. 90X, and from a comparison of the two sketches Beethoven's probable train of thought can be deduced.

The compositional problems posed in the first sketch could be solved as follows. The slow passage at the beginning of the recapitulation in the first sketch could be anticipated, but still remain unexpected, by introducing only a fragment of it at the opening; the problem of

'key structure could be solved, while keeping the major-key element, by using a dominant chord, A major, instead of the tonic; and the slow sections could be related to the allegro by transferring the 'Mannheim rocket' motif itself to the slow part. This last change was surely Beethoven's master-stroke, for by it the theme becomes completely transfigured, now veiled and submerged, beginning deep in the bass clef. The answering phrase, though showing the same descending contour as in the first sketch, is now more animated, with an idea foreshadowed by an isolated sketch on f. 89V/15 (Ex. 3) aind by a rejected idea from Op. 30 No. 2 (f. 557r9). Meanwhile the introduction of recitative into the recapitulation was a way of prolonging the slow passage with a logical continuation of the main four-note motif.

Ex.3

All these changes alter the dramatic balance of the movement, and force other changes. There was by this time a series of interruptions to the forward drive of the movement, and a lyrical, relatively static second subject in B flat would have resulted in too many such interruptions, as well as too many references to major keys. Thus it was necessary to scrap the original second subject, keenino the kev of B flat for the slow movement, and to replace it

" See Kramer, op. cit., pp. 402 & 428, and Reti, op. cit., p. 204.

268

with a non-distinctive, rather unlyrical motif in A minor in the same mood and character as the main part of the movement. In the final version the second subject can hardly be called a theme at all, and it is significant that Beethoven makes no attempt to sketch a second subject on f. 65' of the sketchbook. He simply notes 'A moll erster Theil', and he actually sketches a compressed version of an entire recapitulation without any sign of a second subject. The extended 6/8 coda, by now another unwanted interruption to the rhythmic drive of the movement, has also disappeared, replaced by just a few bars of the tonic chord.

Thus the sketch on f. 90' of the sketchbook can be seen as the true origin of the D minor sonata, and the one on f. 65v as a replacement for it, in which all the compositional problems posed have been resolved. The basic character of the movement is still the same, as is the shape of the first four notes of the main theme; even the layout of the sketch, in the form of a telescoped draft, is retained. But the details have been rearranged and the interrup- tions to the moto perpetuo redistributed, so that the revised version is almost unrecognizable as the same piece. In fact the only passage that remained unchanged for the second sketch was the end of the development with its repeated minim A's and its implied I-V-I-V harmony (Ex. 4)-and even this appears altered and at half speed in the final version (bars 134-7).

Ex.4

etc. I

As the sketchbook was virtually full by the time Beethoven entered the sketch on f. 65', further sketching for the movement would have had to be done on loose leaves, for the next sketchbook, the 'Wielhorsky', was not brought into use immediately. There is now no trace of any such sketches, but one can deduce the direction in which Beethoven would have been working while making them. The bass-line, which in the first sketch tended to fall chromatically, tends to rise chromatically in the second sketch, in both develop- ment and recapitulation: this idea would automatically be transfer- red to the exposition too, as would the working-out of the initial four-note motif. The size of the movement, which in both sketches is quite short, would as usual be expanded considerably (although it remained a fairly compressed movement), with the expansion occurring chiefly in the A minor part of the exposition-the one part of the movement left completely blank in the second sketch. There was also a progressive reduction in the amount of music in

269

the major key. In the first sketch major keys are quite prominent: B flat is used for the second subject, D for the 'dolce' passage, and the first part of the coda goes to F. By the second sketch, B flat major has transferred to the second movement, while the slow passages, though making use of major chords, treat these as dominants of minor keys, and there is no longer any big major section in the entire movement; but bars 7-10 are still fairly confidently in F major, and there is a tierce de Picardie at the end of the movement. In the final version, however, this has been removed, and even bars 7-10 never settle in F major: the introduction of an a'I on the second quaver of bar 9 suggests F minor instead, and although the 'b turns out to be a g'g, rising to at, this has implications of A

minor rather than F major, with the F major chord acting as a submediant. The final version of the movement is noteworthy for its almost complete absence of established major tonality, just as the second movement has an absence of established minor keys.

The next sketches to be considered are those on f. 66' (see Plate I). As stated earlier, it can be surmised that staves 1-2, containing the theme for the finale of Op. 30 No. 1, were filled up when Beethoven was working on the Op. 30 sonatas, but that the rest remained blank until after the insertion of the sketch for Op. 31 No. 2 on the opposite page. Previous writers have regarded these sketches on f. 66' as just a miscellany of ideas for keyboard pieces, connected with no particular work. But it must be remembered that when Beethoven began sketching a multi-movement work he would often, despite starting with the first movement, jot down ideas for later movements at a relatively early stage. Thus concept sketches for all the movements are frequently found before the detailed work on the first movement; presumably he felt it unwise to proceed too far with the opening movement until he had at least some idea of what was to follow. This approach can be seen, for example, in other parts of the 'Kessler' Sketchbook itself, in Op. 30 No. 1 and Op. 31 No. 1.

When viewed in this light, the sketches on f. 66W make perfect sense, for not only are they all in keys that might appear in later movements of a D mninor sonata, i.e. B flat major, D major and D minor itself, but they also show other features that suggest later movements of this sonata. Tlrhe first sketch (Ex. 5), in 3/4 time in B

Ex.5

.-Fr) r ___6

270

-e C C

C

Q

-e

-C

Q Q

0

t C C

C

N I

ji, .... - I C

-C C C-

271

4Al p_F I ' I T

flat, has been described as a minuet,'4 and indeed the words 'alla menuetto' do appear beside it. But this is surely no minuet-least of all a fast Beethovenian one-for its opening theme, bounding unevenly up the keyboard, would sound distinctly comical, if not ludicrous, played in minuet time. The explanation for the words 'alla menuetto' lies in the fact that they are situated at the end of the line: the B flat sketch is for a slow movement, which is to be immediately followed by a minuet. The connections between the sketch and the slow movement of Op. 31 No. 2 are in fact numerous and obvious: both begin with a B flat major arpeggio; both consist of dialogue between low and high pitches; in both cases the low pitches are essentially chordal while the high ones use conjunct melodic motion; and both have alternating tonic and dominant harmony. There can be no doubt that this sketch was intended for Op. 31 No. 2. But it also shows a remarkably close relationship to the first-movement sketch on f. 90v quoted in Ex. 1: both sketches begin with a six-note arpeggio starting on the tonic, followed by a descending phrase that ends with an appoggiatura on to the leading note in bar 4 or 5, so that the sketch on f. 66'- could be described as a major-key variant of the other theme. The second phrase in Ex. 5 is also almost identical to the beginning of the second subject (in the same key) of Ex. 1, indicating that Beethoven intended to transfer not only the key but some of the thematic material from the second subject to the slow movement. Thus Beethoven planned from the outset to make the slow movement very closely related to the first movement in its opening melodic shape, and despite the changes later made to both, the close relationship remained. Out of over a hundred piano sonata movements, these are the only two that begin with an arpeggiando chord, and another obvious relation between the opening themes is the descent to an ornamented feminine cadence on the dominant (bars 6 or 8) at the end of their first sentence.

The remaining sketches on f. 66' seem from details of their layout to have been written down in the order in which they appear on the page, but they do not show the movements in the right order and were probably not written at a single sitting, since they are in varied styles of handwriting. Staves 6-7 represent the start of the minuet announced on stave 3: staves 8-9 show another idea for the

14 Brandenburg, op. cit., p. vii.

272

Ex.6

I r ~. i/ R ; etc.! Adagio; staves 10-11 contain two sketches in D minor presumably intended for the finale; and staves 12-14 contain yet further ideas for the Adagio, this time somewhat more developed than before.

The very short sketch on staves 8-9 is labelled 'adgo' and is in 3/4 time but otherwise shows no very close relationship to any other sketch or to the final version; but the other Adagio sketch, beginning on stave 12, is of much greater interest (Lx. 6). It is still in slow triple time, and although now notated in 6/8, with each bar corresponding to two of the final version, it shows considerable progress over the first sketch. The first two bars are already close to the first four of the final version, especially in the high-pitched answering motifs, now rising instead of falling. The second of these motifs uses the notes c"'-d"'-e"' (with a deleted a" upbeat), as in bar 12 of the final version, rather than a"-b"-c"' as in bar 4, so that bar 4 can be seen as a relatively late alteration to the bar 12 version, rather than the latter being a revision of the former. Indeed Beethoven may even have reached the version in bar 4 by miscopying the number of leger lines and then deliberately deciding to exploit his mistake. Bar 3 of the sketch originally had a diagonal line between its first and second notes. Such lines were sometimoteused in Beethoven sketches as short-hand for a rapid scale, and this is no doubt what is intended here, since the final version has just such a scale at the corresponding place (bar 5). But the diagonal line is one of several features in these bars that have been deleted in the sketch.

In the final version of the Adagio the opening section consists of two sentences: the first (bars 1-8) ends with a half close and the second (bars 9-17), which is a variant of the first, with a full close. In the sketch there is only one statement, with a full close, but this sort of expansion between sketch and final version is quite a common occurrence and was probably intended froe, the start. The first cadence in the sketch therefore corresponds to bars 15-17 of the final version, and is once again close to it, though rather more plain. In bar 5 of the sketch the dotted (or double-dotted) motif is transferred to the bass clef, as in bar 9 of the final version, but after

273

this there is little relation between the two versions. The sketch continues with parallel thirds in the bass clef, and

beneath these, on stave 13, is some sort of alternative, also in the bass clef. At this point a trochaic rhythm begins, and is continued at the beginning of stave 14. Harmonically, however, stave 14 does not quite join on to the end of stave 12 or 13-at least not in Beethoven's 1802 style-and so some transitional music must have been omitted. Stave 14 probably represents the second subject and is in G major. This is not as surprising as it might seem at first sight. Beethoven had just rejected the idea of going to the sub- mediant in the first movement and might well reconsider it for the second; and his next piano sonata movement in B flat-the first movement of the 'Hammerklavier', Op. 106-does go to G major for the second subject, as does the first movement of the 'Archduke' Trio, Op. 97.

No more sketches for this Adagio survive, but there is in the 'Kessler' Sketchbook one idea, intended for another work, that contains related material: this is the deleted sketch for part of the slow movement of Op. 31 No. 1, found at the top of f. 96v (Ex. 7). The figuration of this is very reminiscent of that in bars 51-56 of the slow movement of Op. 31 No. 2, and it draws attention to the similarity between these two adagio movements, both being in a very slow triple time, with simple themes capable of much ornamentation and starting with two bars of tonic harmony followed by two bars of dominant. When Beethoven deleted the sketch on f. 96v he must have felt that there was nothing wrong with the underlying idea but that it was just not appropriate to Op. 31 No. 1; it could still be used later, whereas if he had retained it in this movement it is unlikely that he would have used such a similar idea in Op. 31 No. 2 as well.

Ex.7

:ft t .1 sttl*^ ^ | t: /~~~~~~~~~~~etcl = r =1S~~~~~~~~~~~~~~~~~ =1-X b S

The remaining sketches on f. 66' reveal that Beethoven initially envisaged Op. 31 No. 2 as a four-movement sonata, like the majority of his earlier piano sonatas (nine out of fifteen before Op. 31). In this four-movement plan, the key scheme was to be similar to the recently completed C minor violin sonata, Op. 30 No. 2, with the slow movement in the submediant and the minuet in the opposite mode to the outer movements of the sonata. As in the first two movements, Beethoven initially intended the minuet to mod- ulate to an unexpected key-this time the mediant minor (Ex. 8).

The finale sketches on f. 66' are much less well defined. In the first, Beethoven seems to be planning a fast 2/4 movement in a similar mood and time to the finale of Op. 57; the actual melodic

274

material is similar too, and so yet again the 'Appassionata' was to provide the realization of an idea originating in sketches for Op. 31 No. 2. The second sketch on f. 66/I 1 is even more undistinguished; it seems to be intended for the same movement but in 2/2, with note values doubled and with the main motif of the other sketch in a kind of varied inversion.

Ex.8

IF~ -2JJ If IN 1 I 2

- I~~~~~~~~~~~~~~~~~~I

t$; [S] ' J J J

2_ ! !2_ ' I

Clearly Beethoven had at this stage no fixed idea of what sort of finale the D minor sonata was to have. Eventually, however, he was to build on an idea that had occurred to him sporadically throughout the 'Kessler' Sketchbook. The idea appears in the sketchbook in various keys and guises, and its essence is a theme beginning with a series of repeated notes (usually the tonic) on strong beats, each preceded by some kind of anacrusis, and usually with some arpeggiated accompaniment; but the details vary with each appearance. The first appearance in the sketchbook is on f. 3 , where Beethoven already seems to have the idea of writing a piano sonata in a minor key-here A minor. There is a sixteen-bar sketch (staves 5-8) for a 2/4 Presto, an idea in A major (staves 13-14) probably intended as a slow movement, and between these two (staves 9-12) an Allegro in A minor that foreshadows the finale of Op. 31 No. 2 (Ex. 9). The next similar idea (f. 68'/3-4) is a curious eight-bar theme in E flat, complete with pedal markings, repeats and even some fingering; the metre is now 3/4 and the shape of the anacrusis is identical to that in the D minor sonata. The sketch is quoted in full by Nottebohm and is described by Tovey as 'an

275

Ex.9 a moll

f' Ii IN LjI

allegrl: etc.2 $

absurd danza tedesca'.'5 A third related idea is a four-bar fragment beginning in C major (f. 91v/5-6): this retains the 3/4 metre but alters the shape of the anacrusis; it does, however, use an accompaniment figure almost identical to that in the finale of Op. 31 No. 2. The fourth idea (f. 95r/15-16), also quoted by Nottebohm,'6 was specifically intended for the second sonata of the set, but this is here in A minor and in 6/8 time; the same shape of anacrusis is present, but there is an additional appoggiatura on the strong beat. Some elements from these four sketches are also found in the D major minuet sketch cited earlier, indicating that Beethoven still intended to use the idea in the second sonata of the set, even if in D major rather than A minor and in the minuet instead of the first movement.

The final solution saw a sixth version of the idea, now in D minor for the first time, in a movement that superseded both minuet and finale. The decision to abandon the minuet must have come by the time that Beethoven fixed on the new opening for the finale, for it would have been unthinkable to have a finale so similar rhythmically to the minuet. There was also the precedent of Op. 14 No. 2: this was his only previous sonata with a 3/8 finale, and it too, lacks a minuet, the finale itself being the 'scherzo' instead. As for the main theme of the finale of Op. 31 No. 2, there is a tradition going back to Czerny that the theme occurred to Beethoven when he saw someone riding past his window at Heiligenstadt.'7 Such anecdotes are notoriously unreliable, and the accuracy of this one is certainly called into question by the evidence of the sketches. Did the rider pass by before any of the relevant sketches, in which case it was almost certainly before Beethoven moved to Heiligenstadt? Or were there six riders, one for each variant of the theme? Or did the rider just happen to go past as Beethoven decided to use the theme for the finale of the sonata? At all events it is clear that the movement should not be regarded as a musical representation of a man on horseback.

Sketch study can throw light not only on quaint anecdotes but on analytical and interpretative matters too. It has been apparent

'3 Nottebohm, op. cit., p. 27; D. F. Tovey, notes to Beethoven Sonatas for Pianoforte, ed. Harold Craxton, London, 1931, ii.125.

16 op. cit., p. 35.

17 Ibid., p. 42.

276

ever since Tovey published his performance instructions for Beethoven's sonatas that sketch study can affect modes of performance:'8 although all performers add to the music their own expression, based on their conscious or subconscious understand- ing of the function of individual notes or groups of notes, sketch study often reveals something about Beethoven's understanding of the music, and a performance based on this is likely to communi- cate more of the composer's intentions than one which is not. One place in Op. 31 No. 2 where the sketches hint at Beethoven's understanding of the music is the very first chord, which is not just a chord played arpeggiando but a transfigured arpeggio theme: consequently it should surely be played sufficiently slowly to be heard as such, with extra weight (i.e. prolongation) given to the first note of the arpeggio just as it was in the initial sketch for the start of the movement. The same applies to the start of the second movement, which was also originally a six-note arpeggio theme with a relatively long first note, and which should in performance be made to match the start of the first movement.

Another detail on which the sketches throw light is the problem of what constitute the themes of the first movement of the sonata. The second subject seems not to have been regarded by Beethoven as very important, for it is completely ignored in the sketch on f. 65v; analysts who have denied that the movement has a proper second subject are not far from the truth. The question of whether the first subject occurs at bar 1 or bar 21 has also been much discussed by analysts,'9 and the sketches provide an interesting answer. Bar 21 contains the original first subject as found in the first sketch (f. 90v), but this theme disappears in the second sketch, supplanted by a ,new, albeit related, first subject. Performers wishing to penetrate to the heart of the music by revealing something of the creative process behind it would be justified in emphasising bar 21 at its first hearing as being the original first subject, but passing through it relatively lightly and swiftly in the repeat of the exposition, treating this part of the movement as a mere transitional passage as it is in the second sketch. Such a seemingly eccentric interpretation would not only enable the listener to approach the music from an angle similar to that of the composer; it would also highlight the structural ambiguity of bar 21. Indeed a few pianists already tend towards such an interpreta- tion.

It is now possible to piece together the genesis of the Op. 31 set as a whole. When planning a set of three pieces, whether piano sonatas, violin sonatas, trios or quartets, one of the first things

18 See especially Tovey's notes, ed. cit., ii.51 & 125. 19 Most recently in Martin Wehnert, 'Zum positionellen Aspekt des Thematischen bei

Beethoven', Bericht iuber den internationalen Beethoven-Kongress Berlin 1977, Leipzig, 1978, pp. 335-41.

277

Beethoven had to decide on was his choice of keys, and he seems to have been guided by certain principles that would provide maximum variety and contrast: no two pieces in the set should have either the same key-note or the same key-signature; there should be at least one flat key and at least one sharp key; and one of the three should be a minor key. The latter condition is disregarded in Op. 12, and two of the Op. 1 trios share the same key signature, but otherwise these principles are followed consistently in Opp. 1, 2, 9, 10, 12, 30, 31 and 59; they also apply in the three sonatas of Opp. 13-14, and in Op. 27 if it is grouped with either Op. 26 or Op. 28. In the plans for Op. 31 the minor key was to be A minor ('Kessler' ff. 3' and 95') or D minor (ff. 90v and 65v). The first ideas for a major-key sonata (f. 89v/9-16; f. 90r/14-15) indicate C major, which would preclude the original plan for A minor for the other sonata, and it is at this stage that the first big D minor sketch (f. 90v) appears. Then on ff. 91v\-92r C major is replaced by G major, using a motif originally sketched for string quartet (f. 88r), but incorporating a rhythm and character somewhat similar to those in the C major sketches. The G major ideas are immediately subjected to detailed sketching such as is usually only found for works that were later completed, and so it seems that Beethoven decided straightaway that this was to be Sonata No. 1 of the set. The adoption of G major released A minor once again as a possibility, which is why a 'Sonata 2da' sketch appears in this key on f. 95r before Beethoven returned to D minor. Meanwhile C major, though temporarily abandoned, was to be taken up in the next big piano sonata-the 'Waldstein', Op. 53.

The key of the third sonata, E flat major, was fixed even before Beethoven had decided on the order of the three sonatas, as is evidenced by the sketches in that key on ff. 93r and 95v: both show affinities with Op. 31 No. 3, the former recalling a motif in the second movement and the latter the metre and mood of the first, but both are labelled 'Sonata 2'. Beethoven seems to have attached no special significance to the order of the pieces in a set such as this; and although there are exceptions (such as the Op. 18 quartets), the published order usually followed the order in which the pieces were composed. This is certainly true of Op. 31; but it is important to note that even though No. 3 of this set was not published until about a year after Nos. 1 and 2, the set was conceived from the start as a set of three.

Most of the conclusions presented here cannot be rigorously proved, for there will always be an element of conjecture in this type of study of Beethoven's sketches; but the sketches for Op. 31 No. 2, despite their very limited number, do suggest surprisingly many new insights into the music. They show how the sonata originated in a haze of different ideas which gradually came

278

together in Beethoven's mind into a coherent whole; they also show what features of the sonata were present from the beginning, how these elements were in some cases radically transformed, and what features were introduced only later; and they demonstrate how the very early sketches for a work can be among the most significant for the work as a whole, whereas later sketches would tend to be more concerned with working out relatively minor details. Another interesting feature to emerge is Beethoven's method of transferring ideas, whether of melody, rhythm, figuration, form or key-scheme, from rejected sketches for one work to the finished product of another. This procedure still remains to be explored in detail in the rest of Beethoven's output, but the D minor sonata seems to be related in this way to about half a dozen of Beethoven's works. A full survey of such cross-references would no doubt bring together many more works and would help to highlight those stylistic traits that persisted throughout his career, despite apparent changes of style.

The sketches examined also indicate the need for a redefinition of what precisely constitutes a sketch for a particular work. Beethoven's final sketches for a work often overlap with the writing out of the autograph score, and Lewis Lockwood has drawn attention to the difficulties of trying to draw a clear dividing line between sketch and fair copy.20 There are similar difficulties at the beginning of the sketching process too, so that it is sometimes hard to determine whether certain sketches are intended for a particular work or not. It would be possible to group sketches into four different categories:

(i) sketches clearly related to only a single finished work; (ii) sketches intended for one work but containing ideas developed

in another; (iii) sketches intended for one work but not closely related to its

final form; (iv) sketches apparently unrelated to any finished work.

Hitherto, sketches have generally been regarded as belonging to the first or last types, but there must be a considerable number of sketches, including most of those discussed above, that really belong to the other two. Even this suggested classification is inadequate: further subdivision of these four types would be possible, and no clear dividing-line can be drawn between the first and third categories, since sketches exist over the entire range of possibilities from those that are completely unrelated to any work, through those that nre tn n cyrr-fiar e-r -rvtnt r1i-ntAi-ri to-

2"'On Beethoven's Sketches and Autographs: some Problems of Definition and Interpretation', Acta musicologica, xlii (1970), 32-47.

279

up to those that coincide exactly with an autograph score. But some such classification, which takes into account sketches only distantly related to known works, is surely preferable to the method of listing sketches simply by which work, if any, they closely resemble.

Thus the sketches for Op. 31 No. 2, besides throwing light on its origins and structure and suggesting certain modes of interpreta- tion, can also tell us something about Beethoven's sketches in general, how they should be approached and what problems still have to be faced in their study. Moreover the sketch leaves examined here constitute only a tiny fraction of the number that survive altogether, and they appear in a sketchbook that is comparatively well known and investigated. Much more must still remain to be discovered about Beethoven's music from the numerous lesser-known sketches that survive.

280

Article Contentsp. 261p. 262p. 263p. 264p. 265p. 266p. 267p. 268p. 269p. 270p. 271p. 272p. 273p. 274p. 275p. 276p. 277p. 278p. 279p. 280

Issue Table of ContentsMusic & Letters, Vol. 62, No. 3/4 (Jul. - Oct., 1981), pp. 249-484Front MatterThe New Grove [pp. 249-260]The Origins of Beethoven's D Minor Sonata Op. 31 No. 2 [pp. 261-280]The Sources of 'Cos fan tutte': A Reappraisal [pp. 281-294]The Origins of John Day's 'Certaine Notes' [pp. 295-299]Ivor Gurney's Mental Illness [pp. 300-309]Berlioz, Cassandra, and the French Operatic Tradition [pp. 310-317]Prince Henry as Absalom in David's Lamentations [pp. 318-330]Ariosti's London Years, 1716-29 [pp. 331-351]Contemporary Opinions of Thomas Weelkes [pp. 352-353]Reviews of BooksReview: untitled [pp. 354-355]Review: untitled [pp. 355-357]Review: untitled [pp. 357-361]Review: untitled [pp. 361-364]Review: untitled [pp. 364-366]Review: untitled [pp. 366-368]Review: untitled [pp. 368-370]Review: untitled [pp. 370-371]Review: untitled [pp. 371-374]Review: untitled [pp. 374-376]Review: untitled [pp. 377-378]Review: untitled [p. 378]Review: untitled [pp. 378-379]Review: untitled [pp. 380-382]Review: untitled [pp. 382-383]Review: untitled [pp. 384-385]Review: untitled [pp. 385-391]Review: untitled [pp. 391-393]Review: Opera Guides [pp. 393-395]Review: untitled [pp. 395-397]Review: untitled [pp. 397-400]Review: untitled [p. 400]Review: untitled [pp. 400-402]Review: untitled [pp. 402-403]Review: untitled [pp. 404-406]Review: untitled [pp. 406-409]Review: untitled [pp. 409-412]Review: untitled [pp. 412-414]Review: untitled [pp. 414-416]Review: untitled [pp. 416-417]Review: untitled [p. 417]Review: untitled [pp. 417-420]Review: untitled [pp. 420-423]Review: untitled [pp. 423-425]Review: untitled [pp. 425-428]Review: untitled [pp. 428-430]Review: untitled [pp. 430-432]Review: untitled [p. 432]Review: untitled [pp. 432-434]Review: untitled [pp. 434-436]Review: untitled [pp. 436-437]Review: untitled [pp. 437-440]Review: untitled [pp. 440-442]Review: untitled [pp. 442-443]

Reviews of MusicReview: untitled [pp. 444-445]Review: untitled [pp. 445-446]Review: untitled [pp. 446-447]Review: untitled [p. 448]Review: untitled [pp. 448-450]Review: untitled [pp. 450-451]Review: untitled [pp. 451-452]Review: untitled [pp. 452-454]Review: untitled [pp. 454-455]Review: untitled [pp. 455-457]Review: English Songs [pp. 457-458]Review: French Chansons [pp. 459-461]Review: untitled [pp. 461-466]Review: untitled [pp. 466-470]Review: untitled [pp. 470-471]Review: untitled [p. 472]Review: untitled [pp. 472-474]Review: untitled [pp. 474-475]Review: untitled [pp. 475-476]Review: untitled [pp. 476-477]Review: untitled [pp. 477-478]Review: Recent Polish Music [pp. 478-481]

CorrespondenceFuenllana's 'Orphenica Lyra' [pp. 481-482]A Guide to Electronic Music [p. 482]

Books Received [pp. 483-484]Back Matter