BPM home | Volume 7 index | Article

Fingering as compositional process: Stravinsky's Sonata sketchbook revisited

Graham Griffiths


The composer composes. The insertion of fingering into a musical score is traditionally the concern of the publisher who, occupying the twilight zone between art and commerce, will call upon the expertise of an editor – normally an eminent performer, sometimes a musicologist – to perform this specialised, if comparatively menial service. Those with a creative role in society on the other hand (or so Stefan George would have us believe2) breathe air von anderen Planeten, which would explain why Robert Craft, writing in 1992, should declare in a tone of cautious surprise that:

to compare the sketch and published score of the ... Sonata is to discover that piano fingerings were a part of the composing process, though Stravinsky might remove them later as a builder does his scaffolding.3

As initial response to this assertion this article can confirm, from a study of material held at the Paul Sacher Foundation in Basle, that fingerings abound in Stravinsky's sketches, fair copy and personally corrected proofs of this work.4

Yet it is wise to issue a caveat. There is no reason why the presence of fingering in the sketches of a piano work should prove, necessarily, that this is "part of [Stravinsky's] composing process." He may have jotted down some fingering for an awkward passage in order to prevent his concentration from being broken by digital ineptness or hesitation. My task, therefore, has been to examine this 'scaffolding' and ascertain if it is also, in fact, part of the edifice – much as the plumbing and electrical conduits exposed on the walls and ceilings of the Centre Georges Pompidou (in Place Igor Strawinsky) form an intrinsic element of its acclaimed architectural vision. If this were indeed the case, then one would need to take Stravinsky's fingering into account analytically.

The Sonata was composed over eleven weeks between 4 August and 21 October 1924, these being the dates inscribed on the first and last pages of the sketches; 21 October is also the date on which Stravinsky wrote out his fair copy.5 I wish to examine just one of the thirteen musical ideas to which Stravinsky appended specific fingering at the moment of their first annotation. However my analysis begins not on 4 August, but four months earlier, on 13 April, when Stravinsky took time out from the urgent business of completing his Concerto for Piano and Wind Instruments (finished a week later and given its first performance within a month) in order to write down the following opening to a Sonata movement he subtitled 'Preludio'6 – of which Example 1 is the right hand page:

Musical example

Example 1: Sketch 'SONATA – Preludio' (13 April, 1924)7

This initial sketch is remarkable for many reasons, not least for so accurately anticipating the finished score. Some crossings-out notwithstanding, Stravinsky's hand is so neat, so sure that one may be forgiven for mistaking this for the fair copy. The autograph in ink is Stravinsky's original script; it consists of three pages, comprising thirty-two bars of music. Clearly visible towards the end of the third system is some fingering in the left hand, in ink. (The fainter writing in pencil, including fingering for the right hand, is the work of later revision, probably from early August).

It is surprising, as Craft writes, that fingering should be present at all in sketches – even more so in this autograph that pre-dates the sketches proper by four months. These are contexts, surely, when one imagines the composer to be solely concerned with musical composition – with 'getting the notes onto paper.' Stravinsky's purpose here is not editorial for he does not append fingerings – as would an editor – to suggest to others a solution to a difficult passage. The first two lines of the 'Preludio' certainly do present difficulties – for the sight-reader and performer alike – and in response to this the editor engaged for the first edition of the Sonata, Albert Spalding, judged these opening bars to merit no less than sixty-four fingerings. The composer is no editor and offers none.8 Instead, Stravinsky is concerned (in line three) with the arpeggio accompaniment in the left hand for which he indicates a fingering – 5,4,2,1 – to a figure that presents comparatively little technical challenge; nor would any self-respecting pianist require that this fingering be indicated five times in succession – once would surely suffice. Not unexpectedly, Spalding does not judge this passage to require fingering at all.

Looking ahead, the cadence (lower line, bar 3) might suggest Stravinsky's role to be editorial after all. Upon closer examination, however, one realises that his fingering of 1,2,3 for the left hand – far from facilitating performance – makes for a particularly wide stretch of a fifth between thumb and index finger where more practical alternatives are readily available. Yet by indicating this fingering, Stravinsky is committed to revert to the already favoured 5,4,2,1 hand-position for the final arpeggio:

Musical example

Example 2: Sketch 'SONATA – Preludio' (13 April, 1924) – left hand page

However, any suggestion that these markings are indicative of creative intent would need to be supported by evidence that this fingering accompanies its associated musical idea at every stage of the compositional process just as consistently as the notes themselves. Example 3 illustrates how Stravinsky fingers the left-hand passage in Example 1 on its year-long journey from initial sketch to publication in the summer of 1925 – initially by Edition Russe de Musique and subsequently by Hawkes & Son. One may observe how the 5,4,2,1 fingering was consistently indicated for this figure – by Stravinsky, if not by the publishers:

Musical example Musical example

Example 3a–f: Tracing the left-hand passage originally identified with 5,4,2,1 fingering in the 'Preludio' sketch (Example 1) through to the publication of (the second edition of) the Sonata by Hawkes & Son Ltd in the summer of 1925.
Example 3f © 1925 by Hawkes & Son (London) Ltd. Reproduced by kind permission of Boosey & Hawkes Music Publishers Ltd

Stravinsky, then, did not dispense with his fingerings 'as a builder does his scaffolding' – neither here, nor in another twelve passages which can, likewise, be traced with comparable fidelity through all compositional and editorial stages until, as we have seen, Stravinsky loses his heavyweight bout with the publishers in the final round. Whatever his reason, it was important to Stravinsky that this material be always played in this exact manner. The 5,4,2,1 fingering is clearly identified with this accompaniment figure at its conception and it remained central to Stravinsky's experience of this material in performance. What, then, within the wider Stravinskian context, could be the genesis for this uncommon association?

The autograph material for Les Cinq Doigts of four years earlier (1920/1921) does contain fingering of an editorial and appropriately didactic nature: at the head of every movement and principal section Stravinsky even indicates the exact hand position and its location on the keyboard.10

Between this and the Sonata there may be a bridge that links the clearly pedagogic intent of 'The Five Fingers' to the, as yet, ill-defined (possibly creative) process in the later work. A transitional, certainly extraordinary example of Stravinskian fingering may be observed in the fair copy of the Concerto for Piano and Wind Instruments (1924). Stravinsky appears to be engaged here (Example 4) in a concentrated outburst of meticulous fingering annotations. As an isolated manifestation, in a score almost devoid of markings of any kind, it is quite remarkable. One may observe how the 5,4,2,1 pattern (here in the right hand) is marked eight times in succession.11 In fact, the previous ten bars although un-fingered (five are visible on this page) also employ this identical hand-position a further twenty times consecutively:

Musical example

Example 4: Concerto for Piano and Wind Instruments, 1st movement, Fig. 36–7
© 1924 by Hawkes & Son (London) Ltd, revised version © 1960 by Hawkes & Son (London) Ltd. US Copyright renewed. Reproduced by kind permission of Boosey & Hawkes Ltd.

A particularly intriguing detail is to be observed in the passage that Stravinsky has marked: this contains uncomfortably wide leaps between the 2nd and 5th fingers of more than an octave (Example 4: one bar before fig. 37), but these too are made to conform to the one hand position – despite the extreme awkwardness of this at speed, and even though alternatives are available.13 Although enodowed with unusually large hands (as memorably observed by Picasso) Stravinsky recognizes the danger of these out-size stretches and proposes an effective, if simple diagrammatic symbol – a diagonal line – as warning. At all costs, it seems, the pattern must be maintained even when the pianist is more likely to play the correct notes by finding his or her own solution. Fingering is no longer merely the means to a musical end, it seems; here, the music has become a means to a technical objective. By devising a passage of twenty-eight consecutive applications of the same (essentially uncomfortable) hand-position, Stravinsky might be accused of enjoying the composition of technical exercises with the zeal of Clementi, Czerny, Henselt, Kalkbrenner, and Leschetizky; or, even closer to his own experience, of Isidor Philipp with whom he had been studying privately, in Paris, to prepare for his debut as soloist in the Piano Concerto.

Perhaps Stravinsky's effort to compose a soloist's part worthy of the genre led him to resource the available literature of piano tutors and studies. He cannot have been insensitive to the prospect that comparisons would be drawn between his new concerto and those of fellow Russian composer/pianists Prokofiev and Rachmaninov. One might conjecture that, consistent with his recent aesthetical pronouncements – 'Some Ideas About My Octuor' was published in January 1924 – Stravinsky now sought a pianistic idiom free of nineteenth-century rhetoric while guaranteeing an equivalent degree of technical difficulty. His aim, perhaps, was to mould the soloist's material not on the flashing octave displays and massive chords of the virtuosic showman but on the contrapuntal rigour of piano tutors and etudes, on the literature of the musical draftsman immersed in the intimate technical challenges of piano study, persistent and proud of his feats of digital endurance. But this is conjecture. Without additional and substantive evidence to the contrary one must continue to view this 'behaviour' merely as a composer's passing obsession with digital patterns for his own sound, professional reasons.

The above examples cannot alone amount to evidence of a creative process as such. The opening to the final movement of the Sonata, however, may serve as a more reliable witness and indicate the direction that future research might take. Here (Example 5) we find a two-part fugue whose subject is born of that fundamental 5,4,2,1 left-hand position. There is no need to add fingerings this time; the fingers have 'got their way' and Stravinsky has promoted his digital aerobics from accompaniment and passage-work to primary thematic material. Furthermore, not only does the movement unfold musically – with a sense of thematic/harmonic evolution and textural contrast – there is also a technical evolution. If the fugue subject initially resembles an 'exercise for the left hand', this soon leads to material that ratchets up what Soulima Stravinsky referred to as 'technicity'14 by exploiting an additional challenge: the rapid overlapping of the fifth finger by the fourth; see Bar 6 (Stravinsky's fingering). There are other moments, too, where this is clearly implied – for example, between bars 2–3 and bars 4–5 (my fingering is in square brackets):

Musical example

Example 5: Sonata, 3rd movement, bars 1–815
Example 6: Bars 71–8
© 1925 by Hawkes & Son (London) Ltd. Reproduced by kind permission of Boosey & Hawkes Music Publishers Ltd

Continuing with the Piano Method analogy, one might then expect to progress to a test of augmented difficulty – intended to develop these overlapping skills – now at the upper extremity of the left hand, between the thumb and second finger. Indeed, this is to be found later in the movement (Bar 72 – with Stravinsky's fingering, and Bar 73 – with my fingering in square brackets), as illustrated in Example 6.16 Both these challenges – the fourth finger overlapping the fifth and the second overlapping the thumb – are later combined in Bars 77–8 (the final two bars of Ex.6).

With reference to the Concerto, Sonata and Serenade Charles Joseph (who assumes inside knowledge, having studied the piano with Soulima) declares 'it is impossible to miss [the] reliance [of these works] on Philipp's exercises as useful models' in order to justify what he refers to as 'Stravinsky's unique compositional approach to the keyboard.' Not only does he assert that entire passages in Capriccio are 'virtually lifted' from Czerny's Op. 337, he also declares that '[o]ne of the great underestimates in Stravinsky scholarship is of just how important these pianistic, tactile models were in shaping the composer's writing during [this] period.' But Joseph has a different agenda and he fails to recognise that the gauntlet he has thrown down is itself a five-finger envelope by another name. Instead, he attributes those pianistic textural figurations in non-pianistic Apollon Musagète merely to 'the anatomy of the composer's unusually large hands.' There is clearly more to it than that.

Like Stravinsky's piano writing, doctoral research is, by its very nature, a feat of endurance both intellectual and digital. Over time this burgeoning project may yet address Kofi Agawu's premise that 'musical composition is re-composition ...[something] simply inconceivable outside a specific pedagogical tradition'17; However Heinrich Schenker, surely voicing the opinion of many, has written:

A musical person, having reached a certain level, does not need to practise in the sense of finger exercises and etudes; such practising only leads back to these very exercises, to these very etudes – a world not worth reaching ... 18

To propose Stravinsky's fingering as an indicator of musical and gestural significance – a determinant, even, of compositional process – carries wide-ranging implications both for the analysis of other piano works and the broader context of Stravinsky scholarship: not least, how such a study might contribute to a re-evaluation of the composer's 'objective' aesthetic and his views on performance. In 'Some Thoughts On My Octuor' Stravinsky had written that he '[did] not conceive nor feel ... true emotive force except under coordinated musical sensations.'19 By November 1924 he was further expounding his novel aesthetical stance, this time through the medium of the piano: less than a month after completing the fair copy of the Sonata, Stravinsky gave the first private performance of this work (in a hotel in Warsaw). His first comment, immediately following the conclusion of the final movement, is reported to have been: '[Y]ou see, I create like this: from colour I go to the drawing. Right now I am struggling to create drawings at the expense of colour.' The whole experience led one critic to comment that '[Stravinsky] used to walk on a rope, then a wire. Now he walks on a razor blade',20 while the response of the poet and novelist Jarosław Iwaszkiewicz is more perceptive:

One is amazed by the unusual logic and mathematical clarity that are apparent in everything [Stravinsky] says, and most of all [in his] objectives and methods. ... Total awareness in all his creative work – that is what Stravinsky develops in himself and what he strives towards.21

A sense of total awareness with undertones of an unusual logic – be it mathematical or monochrome – is indeed suggested by Stravinsky's instrumental counterpoint of this period. In the specific instance of the Piano Sonata, if the available edition of this work were to reproduce Stravinsky's intended fingering – and if pianists were to adhere to this as scrupulously as to the written note – they might also harness those 'coordinated musical sensations' and share the composer's 'emotive force.' Bound by this digital strait-jacket like a human pianola, the performer might also more faithfully grant Stravinsky his wish – fervently expressed in Poetics – that his music be executed, not interpreted.


1 A version of this paper was first presented at the RMA Research Students' Conference in Durham, 30 March–01 April 2005 Back

2 Cf. Schoenberg's George setting in String Quartet No. 2 (1908). Back

3 Craft, Robert: Igor Stravinsky: Glimpses of a Life (London: Lime Tree, 1992), p.329 Back

4 I am indebted to the Paul Sacher Foundation for granting me access to the Stravinsky Collection (November–December, 2004) – in particular to Dr. Ulrich Mosch for his helpful and expert guidance during and since my visit. Sketches of the Piano Sonata are reproduced in this article (Examples 1 and 2) with the kind permission of the Sacher Foundation. Back

5 The fair copy is in a separate, hard-bound book held in folder R 91 at the Paul Sacher Foundation, Basle. Back

6 This sketch, titled 'Preludio', was written into a sketchbook now catalogued at the Paul Sacher Foundation as 'Skizzenbuch VII'. It can be viewed on 'Mikroformregister' 123, frames 0516 and 0517. Back

7 Stravinsky: Sein Nachlass, Sein Bild (Basel: Basel Kunst Museum, 1984), pp.102–103 Back

8 Neither here nor, for this passage at least, at any stage between this sketch and the final publication. Back

9 Stravinsky, Sonata, piano solo, ed. Albert Spalding (London: Hawkes & Son Ltd., 1925) Back

10 All of these details were faithfully reproduced by the publishers J & W Chester (1922) Back

11 Stravinsky's indications begin four bars before figure 37 of the published score (Copyright 1924 by Hawkes & Son, London, Ltd.) Back

12 Igor Stravinsky, Concerto for Piano and Wind Instruments (London: Hawkes & Son Ltd., 1924), 26. Back

13 One such alternative is the repetition of the thumb three bars before fig. 37 Back

14 As recalled by Charles M. Joseph in his book Stravinsky Inside Out (New Haven: Yale University Press, 2001), 91 Back

15 Stravinsky, Sonata, piano solo, ed. Albert Spalding (London: Hawkes & Son Ltd., 1925), 14 (lines 1 and 2), bars 1–8. Back

16 Stravinsky, Sonata, piano solo, ed. Albert Spalding (London: Hawkes & Son Ltd., 1925), 17 (lines 1 and 2), bars 71–8. Back

17 '... the indisputable fact [is that] musical composition is re-composition ... [something] simply inconceivable outside a specific pedagogical tradition.' Kofi Agawu, "Schubert's Sexuality: A Prescription for Analysis?" in Nineteenth-Century Music, 17 (1993), 81 Back

18 Heinrich Schenker, Kunst des Vortrags, ed. Heribert Esser; trans. Irene Schreier Scott as The Art of Performance (Oxford University Press, 2000), 75 (my italics) Back

19 Igor Stravinsky, 'Some Ideas About My Octuor' printed in The Arts, January 1924; reprinted in Eric Walter White: Stravinsky: The Composer and his Works (London: Faber and Faber, 1966), 576 Back

20 As reported by Jarosław Iwaszkiewicz in Wiadomosci Literackie (Warszawa) 1924, no. 46; in Viktor Varunts, Igor Stravinsky: Publitsist i Sobesednik (Moscow: Sovetskiy Kompositor, 1988), 47. Assistance from Tatyana Ursulova in the translation of this text is gratefully acknowledged. Back

21 Jarosław Iwaszkiewicz, Wiadomosci Literackie (Warszawa) 1924, no. 46; in Viktor Varunts, Igor Stravinsky: Publitsist i Sobesednik (Moscow: Sovetskiy Kompositor, 1988), 47 – my italics Back

© Graham Griffiths, 2005

How to cite this article

BPM home | Volume 7 index