The handwriting of any individual is a kind of self-portrait, and reading a handwritten letter can give an indication of the writer’s character and state of mind, and of his or her attitude to the content of the letter. An author’s manuscript often yields valuable information about the creative process; the manuscripts of Dickens or Dostoevsky provide many examples. Examining such a document is a very different experience from reading a novel in cold print. With a musical manuscript, the spacing, the character of the pen-strokes and of the musical handwriting, as well as details of layout which cannot always be exactly reproduced by the process of engraving, give similar information, valuable to the student and to the performer. Beyond factual information, the visual impression of the manuscript, the Notenbild, can be a direct stimulus from the composer to the interpreter’s imagination. In this way, study of the composer’s manuscript can lead both to a narrowing of the possibility of textual error and to a widening of the possibilities of imaginative response to interpretation.
Examining the manuscripts of any great composer or literary author is always a thrilling experience. I have spent many hours studying Scriabin’s manuscripts in the Glinka Museum, Moscow, which holds in its vast archive many fair and rough copies of complete works as well as sketches by Scriabin. The first thing which strikes one is the extreme beauty and clarity of the scores. The slender exactitude of the writing and drawing corresponds to the delicacy and transparency of Scriabin’s own playing of his music, and makes it clear to the interpreter that a similar clarity, precision and grace is demanded in his or her own performance – something extremely difficult to achieve. The care with which the manuscripts were prepared confirms the testimony of Scriabin’s friend and biographer, Leonid Sabaneyev, who was bemused by the care taken by the composer in the placing of slurs, the choice of sharps or flats in accidentals (contributing in many cases to an analysis of the harmony concerned), the spacing of the lines of the musical texture over the staves and the upward or downward direction of note stems.
It was Heinrich Schenker who pointed out the expressive and structural significance of the manuscript notation of Beethoven, and who was instrumental in establishing an archive in the Austrian National Library, Vienna, of photographic reproductions of musical manuscripts. His pioneering work has led gradually to the present wealth of Urtext editions and facsimiles of many composers’ manuscripts. Reproductions of Skryabin’s manuscripts have been published by Muzyka (Moscow), Henle (Munich) and the Juilliard School (New York; their manuscript collection is available online). These reproductions cover several significant compositions by Scriabin: Sonata no. 5, op. 53; Two pieces, op. 59; Poème-nocturne, op. 61; Sonata no. 6, op. 62; Two poèmes, op. 63; Sonata no. 7, op. 64. The remarks below have no pretensions to system or completeness; they are merely observations based on initial study, and intended as a stimulus to others to examine the manuscripts for themselves.
In maturity, Scriabin took immense care with his manuscripts. Speaking to Sabaneyev, he compared the difficulty of writing down a conception in sound to the process of rendering a three-dimensional object on a flat surface. As a student and as a young composer, though, Scriabin was by no means ideally accurate or painstaking in his notation. This was the cause of Rimsky-Korsakov’s irritated response to the manuscript score of Scriabin’s Piano Concerto – the elder composer initially considered it to be too full of mistakes to be worthy of serious attention. Mitrofan Belaieff, Scriabin’s publisher, patron and mentor, frequently begged the composer to be more careful in correcting proofs. The original editions, particularly of the early works, contain many errors which originate in some cases from Scriabin’s manuscript and in others from poor proofreading – as far as we can tell; some early manuscripts are now lost.
We are indebted to the fine musician Nikolai Zhilyayev for correct editions of Scriabin’s music. Zhilyayev knew Scriabin well and discussed many misprints with the composer; others he detected by his own scrupulous and scholarly work and prodigious memory. As Scriabin’s harmony and voice-leading were impeccably systematic and logical at all stages of his development, those who have had to do with the old editions will know that is often possible to correct mistakes by analogy or knowledge of harmonic style.
Zhilyayev was the revising editor for a new edition of Scriabin’s music, published by the Soviet organisation Gosudarstvennoe izdatel’stvo muzykal’nyi sektor (State publisher, musical division – ‘Muzsektor’) from the 1920s on, each work or opus number being issued separately. These beautifully prepared editions are painstakingly annotated, corrections being indicated in two layers: those discussed with the composer and therefore beyond doubt, and those which Zhilyayev considered likely (and he was usually right). This work was the basis of the complete edition of the piano music published by Gosudarstvennoe muszykal’noe izdatel’stvo (State musical publishing house – ‘Muzgiz’) in three volumes (1947, 1948 and 1953). Zhilyayev fell victim to Stalin’s terror; he was arrested in 1937 and shot in the following year. His name does not appear on this three-volume edition.
A new complete edition is appearing gradually under the imprint Muzyka–P. Jurgenson. The general editor is Valentina Rubtsova, biographer of Scriabin and head of research at the Scriabin Museum, Moscow, assisted by Pavel Shatsky. As in Rubtsova’s editions for Henle, full credit is given to Zhilyayev, and the annotations as to origins and variants are very thorough in this valuable new edition.
A very limited number of Scriabin’s manuscripts has been available in facsimile until now. The collection of ‘Youthful and Early Works’ prepared by Donald Garvelmann and published in New York in 1970 by Music Treasure Publications contains a facsimile of the early E flat minor sonata (without opus number) of 1889, a typical youthful manuscript of the composer, rather heavy in its style of penmanship. The manuscript of the op. 11 preludes (excerpts are shown in ill.1), though tidier, shows a similar style.
The Russian website ‘Virtual’nye vystavki’ (‘Virtual exhibitions’) gives in facsimile the first page of the Etude op. 8 no. 12, with more fingering than is shown in the Belaieff edition, and also the first page of the manuscript score of the Poem of Ecstasy, providing a striking example of the change in the composer’s manuscript style. A facsimile on the site of the first two pages of the score of the Piano Concerto shows some of the copious blue-pencilling of Rimsky-Korsakov from the occasion mentioned earlier, and the site also reproduces Skryabin’s letter of apology to Rimsky-Korsakov apologising for the errors and blaming neuralgia. Comparison of this letter with the one to the musicologist N. F. Findeizen dated 26 December 1907, also viewable on the site, gives another clear example of the change in Scriabin’s handwriting.
Op. 53: Sonata no. 5
A facsimile of the Fifth Sonata has been published by Muzyka. The manuscript of this work was presented to the Skryabin Museum, Moscow, by the widow of the pianist and composer Alfred Laliberté, to whom Scriabin had given the manuscript. This is a very different document from the early E flat minor sonata manuscript, and shows Scriabin’s fastidious and calligraphically exquisite mature hand. By this time both Scriabin’s music manuscript and his handwriting had developed an elongated ‘upward-striving’ manner. We might make a comparison with the remark of Boris Pasternak that the composer ‘had trained himself various kinds of sublime lightness and unburdened movement resembling flight’ – the handwriting is expressive of this quality. Examples of Scriabin’s handwriting in letters to Belaieff in 1897 (ill. 2) and to the composer and conductor Felix Blumenfeld in 1906 (ill. 3) show the dramatic difference in handwriting style that developed.
The manuscript of the Fifth Sonata shows that, although Scriabin spoke French, he did not immediately provide a French text for the epigraph, which is from the verse Poem of Ecstasy. This poem was written in Russian at the same period that the symphonic poem was composed. There is a request on the manuscript to the engraver to leave space for a French version. The French text, which is the usual source of English translations, does not reflect the Russian with complete accuracy: the forces mystérieuses, ‘mysterious forces’, which are being called into life are skrytye stremlen’ya, ‘hidden strivings’, in the original. In other words, it is open to doubt that any sort of ‘magical ritual’, in a superstitious sense, is being depicted in this work, a suggestion made (perhaps in a figurative sense) by the early writer on Scriabin Evgenii Gunst and elaborated upon by the composer’s British-American biographer, Alfred Swan. The epigraph may be regarded as an invocation of Scriabin’s own inner aspirations, the creative power which the composer equated with the divine principle.
Work on the Fifth Sonata started in 1907, at a period when a rift had developed between Scriabin and the publishing house of Belaieff. The committee running the publishers after the death of Belaieff had proposed a renegotiation of fees. It is possible that Scriabin was unaware of the preferential and generous treatment Belaieff had accorded him; certainly, he was offended by the proposals and withdrew from his agreement with the publishers. The Sonata was published at Scriabin’s own expense, but was taken into the publishing concern run by the conductor Serge Koussevitsky, Rossiiskoe muzykal’noe izdatel’stvo (RMI). Later still, Scriabin quarrelled with Koussevitsky too, and the composer’s last works were published by the firm which had brought out his very first published compositions, Jurgenson.
The main differences between the manuscript of the Fifth Sonata and most modern printed texts are:
1) a missing set of ties at the barline between bars 98 and 99. These ties are also missing in RMI, and in the edition printed at Scriabin’s own expense. Muzgiz adds the ties in dotted lines, by analogy with the parallel passage at bars 359–360. The commentary to the Muzgiz edition states that sketches of the work make use of an abbreviated notation at this point which could have led to this misunderstanding, as the editors describe it. Christoph Flamm’s notes to the Bärenreiter edition are definite as to Scriabin’s intention not to tie over this barline, citing the repetition of accidentals in bar 99 as being conclusive proof.
2) the movement of the middle voice in bars 122–123, 126–127, 136–137, 383–384, 387–388, 397–398 (Meno vivo sections): the manuscript gives a downward resolution in the middle voice (d flat – c in the first passage and g flat – f in the second) whereas the printed editions give an upward resolution (d flat – d natural and g flat – g). It is as if only at a second attempt (as revised for the printed version) has Scriabin fully realised the implications of his own (then very new) harmony: the resolutions as printed resolve into the augmented harmony around them, whereas the resolutions in the earlier version do not. Knowing about this early version, moreover, adds point to the grandiose version of the same section at bars 315–316, 319–20 and 323–324, where the downward resolution is retained. One might think of the meno vivo sections as being potential states, and the grandiose version as representing a fully realised condition.
It should be remembered that the Sonata was composed at breakneck speed, completed in a few days, and revised afterwards; Valentina Rubtsova, editor of the facsimile, suggests that the manuscript provides a glimpse into the composer’s creative laboratory. She further points out that Scriabin uses double barlines to indicate structural divisions, whereas publishers’ house style often requires a double bar at any change of key-signature or time-signature. This has resulted in the insertion of a number of non-authentic double bars in some published versions of the Fifth Sonata. Double bars occur in the manuscript in the following places only:
before bar 47 (beginning of main sonata exposition)
before bar 120 (Meno vivo, the second subject area)
before bar 367 (indicating, perhaps a slight hesitation before this rising sequence)
before bar 381 (parallel passage to bar 120).
The visual effect of the manuscript is therefore more continuous than that of the printed edition. It should be mentioned that the Urtext printed version given in the volume containing the manuscript is a corrected version of the RMI edition. This edition was prepared with the composer’s agreement and during his lifetime. The manuscript, though, is an invaluable source for the reasons given above.
A similar use of double barlines to that in the Sonata no. 5 is made elsewhere by Scriabin, including in the Sonata no. 6 (see below) and the Sonata no. 8. It can be said, from these examinations, that Scriabin uses double barlines structurally or even expressively, and that they often should be made audible in some way, in sharp contradistinction to the purely ‘grammatical’ double bars referred to above. The definition of ‘sometimes’ and ‘often’ is a non-scientific one and comes down to the player’s own interpretive insight, but where there is a double barline and no change of time- or key-signature, the double bar clearly has structural significance.
The addition of a double bar by a publisher can confuse the interpreter. For example, Bach’s engraved edition of his own Second Partita has no double bar, in fact no barline at all, at the beginning of the third section of the Sinfonia (ill. 4.) The insertion of a double bar at this point, even in some ‘Urtext’ editions (because of the change of time-signature) leads many performers to treat the final chord of the middle section like a ‘starter’s pistol’ for the quicker final section, which, as consideration of the musical content will quickly demonstrate, starts on the second quaver of the bar with the fugue subject.
The notes by Valentina Rubtsova to the facsimile of the Fifth Sonata mention Scriabin’s differing use of rallentando in its full version and of the abbreviation rall., and the possible implications of such usage for performance:
[…] in b. 382 Scriabin indicated molto rallentando, while in b. 386 and 390 he confined himself to [a] shortened and somewhat careless rall. It seems that the theme sounded to him just like that: with a more substantial broadening in b.382 and in a somewhat generalized manner in b. 386 and 390.
A related expressive function of details in the writing of performance directions will be noted below in the case of the Poème-Nocturne, op. 61.
Now we move to a group of Scriabin’s manuscripts, recently published on line by the Juilliard School of New York. The works with opus numbers 52, 53 and 58 to 64 were published by Koussevitsky’s firm, RMI, mentioned above. (The Poem of Ecstasy, op. 54, was already contracted to Belaieff, as were opp. 56 and 57; there is no work with the number 55.) Opp. 59 and 61 to 64 (op. 60 is an orchestral score, Prometheus) were bound together in one volume at some time. Koussevitsky’s archive went with him when he left Russia in 1920. The majority of the archive is now in the Library of Congress, but this volume somehow came onto the open market, and was sold at Sotheby’s in 2000. The purchaser, Bruce Kovner, businessman, collector and philanthropist, generously donated his entire collection to Juilliard School in 2006, and Juilliard have made the contents of the volume he purchased available in excellent facsimile online – a huge step forward in making Scriabin manuscript facsimiles available to the musical public. The Sonata No. 7 has also been published in an equally excellent facsimile by Henle with informative notes by Valentina Rubtsovsa. Some observations on these manuscripts follow.
Op. 59 no. 1, Poème
- b. 13: the two d sharps (the sharp sign was omitted) in the r. h., third and fourth quavers, are not tied. Perhaps they should be made uniform with the beginning, but this small variant could be seen as going with the more significant change of r.h. rhythm in b. 15. This detail raises the familiar problem: to what extent should a manuscript be preferred to a first edition? The editor has not queried this missing tie, but the tie was inserted in the first edition – possibly, a correction at proof stage by Scriabin. Cf. b. 36 for another tie which is missing in the manuscript but present in the first edition, probably corrected by Scriabin.
b. 15: an accidental is missing before the r. h. d sharp, third quaver of the bar. This mistake, as well as the missing accidental in b. 13, was reproduced in the first edition, but corrected by Zhilyayev.
b.19: the fifth quaver in r. h. is spelled in the manuscript as b double flat, harder to read than the a natural printed in most editions, but consistent with the d flat bass of this bar and typical of Scriabin’s fastidiousness in his choice of accidentals. The spelling was reproduced in the first edition, but altered without comment by Zhilyayev, who did not have the manuscript available. (This manuscript was also not available to the editors at the time of preparation of the Muzyka-Jurgenson edition.) Subsequent editions, including Muzyka-Jurgenson, followed Zhilyayev’s reading. The ‘spelling’ of a note may well have an effect on the player of a wind or string instrument as regards actual pitch, and Sabaneyev discussed this with Scriabin. But a good pianist will often respond by minute adjustments of touch to the difference of inner hearing caused by enharmonic differences of spelling.
b. 23–25: there is evidence in these bars of erased octave doublings in the right hand phrases, though the lower octave to the initial a, r.h. second quaver of bar 23, has not been erased – a mistake rightly queried by the editor. Here, the texture is delicate and transparent, but it will be remembered that Scriabin often preferred single notes to octaves in passages of powerful sonority where an effect of brightness was desirable (e.g. final climaxes of the Fifth Sonata and Vers la Flamme). Sabaneyev criticised the composer for scoring his orchestral music with doublings at the unison rather than the octave, but this seems to have been Scriabin’s preference in many places.
b. 28 and 30: The three r.h. quavers which continue the middle voice at the end of these bars were first written by Scriabin in the upper staff, but then erased and put into the lower staff, clarifying the voice-leading. This is an example of the care taken by the composer in the optical presentation of his voices.
b. 34: the manuscript and the first edition have d natural in r.h. upper voice, second, fourth and sixth quavers. This error was corrected by Zhilyayev, who changed these notes to d sharps, noting the analogy in bar 6.
b. 36: the tie between third and fourth quavers of the bar in r.h. is missing in the manuscript, but was supplied in the first edition – possibly a correction in proof by the composer.
b. 38: the acciaccatura at the beginning of the bar for both hands was written by Scriabin with a quaver tail without the customary cross-stroke. This seems to have been the composer’s usual habit – compare the beginning of the Sixth Sonata, written in the same way, as well as other instances – and, in the case of the present Poème, the notation was altered in the first edition. The RMI edition of the Sonata, however, shows the acciaccatura with a quaver-type tail, though many later editions add a cross-stroke. It may be felt that in both cases Scriabin’s notation may suggest a more deliberate execution of the acciaccaturas.
b.39: note the beautiful and unusual notation of the final sonority, a single stem uniting sounds many octaves apart and played by two hands. It is suggestive of the deep and strange sonority of this ending. It is given by most editions, but not by Peters, who ‘normalise’ the notation here.
Op. 59 no. 2, Prelude
A number of errors in the manuscript were correctly questioned by the editor, and further inconsistencies were corrected by Zhilyayev.
The rhythm at the beginning of bar 40, though, (marked avec defi – Scriabin omitted the acute accent on the second letter of défi) written as three even quavers, was retained in the first edition and subsequent ones despite having been questioned by the editor. Muzgiz, following Zhilyayev, queries whether it should be made consistent with the dotted rhythm of other similar bars. The Peters edition by Günter Philipp adopts this suggestion. The present writer is of the opinion that the three even quavers help to express Scriabin’s suggested ‘defiance’.
Intriguingly, a slip of paper was pasted over the original manuscript at bars 26–28. This is at the position, characteristic of Scriabin’s short pieces, where the opening material begins to be repeated in transposition. The repeated chords on the paper slip, which anticipate the coda from bar 54 to 57, may have been a late compositional addition by Scriabin. (Other paper slips are observable, pasted into the manuscript of the Sonata no. 6.)
Op. 61, Poème-Nocturne
(The manuscript of this work was also not available to the editors of Muzyka-Jurgenson, who were, however, able to consult a rough draft, as in the case of op. 59.)
Space will not permit a detailed analysis of longer works such as this, but some interesting features present themselves. The first page of the manuscript is written in two inks, blue and black. On the first system, the clefs and the r.h. phrase from the downbeat of bar one are written in blue, whereas the upbeat is written in black. A list of incipits for projected works by Scriabin exists in the Glinka Museum archives, and has been examined by the present writer. This list corresponds to a description by Sabaneyev of a collection of thematic material ‘for sonatas’. In the list, the Poème-Nocturne theme lacks its upbeat. Perhaps the addition of the upbeat was a late inspiration, like Beethoven’s last-minute addition of a two-note upbeat to the slow movement of the Hammerklavier sonata. At the recapitulation, b. 109, the theme starts on the downbeat.
In bar 3 and the corresponding passage, bar 110, Scriabin writes the molto più vivo directly over the l. h. figure on the second beat. This is placed too late in Muzgiz, but correctly in Muzyka-Jurgenson.
Scriabin’s usual practice is to write his performance directions or remarki in lower-case letters, but in the Poème-Nocturne and some other works this practice is departed from in certain places. The new ideas at bar 29 and 33 are marked in the manuscript Avec langueur and Comme en un rêve – suggesting, perhaps, that the arrival of these new ideas should be ‘shown’ by the player in some way, possibly by a very slight elongation of the rests before them, as with the start of a new sentence or paragraph in a text which is read aloud. The same thing happens at Avec une soudaine langueur (sic) in bar 52, and Avec une passion naissante and De plus en plus passionné in bars 77 and 79. The first edition reproduces this peculiarity, but not Muzgiz or Muzyka-Jurgenson. It has not been possible to determine whether they are following Zhilyayev, as seems likely.
The addition in printed editions, including the first, of a poco acceler. [sic in RMI] over the barline of bb. 46-47 is clear evidence of intervention by the composer at proof stage.
The long slur at comme un murmure confus (bar 103 to 110) is correctly reproduced in the editions known to this writer, but seeing it drawn so clearly and with such certainty in the manuscript is a reminder not to yield to the temptation to ‘explain’ the structure of this mysterious passage, and especially not to render the arrival of the recapitulation in bar 109 with any excessive degree of clarity. The piece reflects Scriabin’s exploration of states of consciousness on the borders of sleep, as he explained to Sabaneyev. On the other hand, the remarka at the point of arrival of the recapitulation (Avec une grace [sic] capricieuse) does have the capital letter we have come to expect in this work when important thematic ideas are presented.
Op. 62: Sonata no. 6
This work is so successfully suggestive of dark areas of the spirit that a listener once suggested to the present writer, after a performance of the Sixth Sonata, that the music was evidence of psychosis in the composer’s own mind. The listener was, of course, making an error like that of Don Quixote at the puppet show – mistaking dramatic presentation for reality. The lucidity of the manuscript, as well as the highly organised and disciplined musical structure, show that Scriabin knew very well what he was doing.
Towards the end of the work there is a notorious high d written, which exceeds the range of the keyboard (bar 365). This note has also been quoted to me by music-lovers as evidence of Scriabin’s supposed delusional condition. Firstly, it should be pointed out that the d is dictated by analogy with bar 330. We can make a comparison with Ravel in this case. In the climax of Ravel’s Jeux d’eau there is a bottom note which, harmony dictates, should be G sharp, but as the note does not exist on most keyboards, Ravel wrote A. Similarly, Ravel ‘faked’ octaves at the bottom of the piano in the recapitulation of Scarbo by writing sevenths. Scriabin, ever an idealist, preferred to write the pitch required by the music and to leave the solution to the interpreter. Furthermore, the whole phrase from bar 365 to 367 is written an octave lower in the manuscript than in the first edition, thus bringing the d within the keyboard range. An explanation for the late change between manuscript and first edition, which transposes the phrase up an octave, may be that Scriabin never performed this very difficult work – the premiere was entrusted to Elena Beckman-Shcherbina. Perhaps, in working on the piece with her and hearing the passage played up to tempo, Scriabin suggested that she try the phrase an octave higher, as the analogy with bar 330 demands, and realised that the chord flashes by with the substitution of c for d as the top note practically unheard. In her memoirs, Bekman-Shcherbina describes Scriabin’s detailed work with her on his compositions, but, alas, gives no details of the work which must have taken place on the Sixth Sonata.
The composer’s notation of the acciaccatura which starts the Sixth Sonata has already been mentioned (see above, Poème op 59 no. 1.) As in the case of the acciaccatura which sets off the Sonata in A minor by Mozart (K.310), this opening should not be played too glibly, but with a certain weight. Indeed, for a player whose hand cannot stretch the initial chords, it is a help to know that this arresting opening should not be hurried over. More importantly, an execution on the slow side helps to emphasise the sombre, unyielding severity of the opening sonority. It is perhaps unfortunate that publishers’ ‘house styles’ lead to a routine ‘correction’ of Scriabin’s notation of the acciaccatura.
‘House style’ has also led to the omission in some editions of the Sixth Sonata of a number of ‘structural’ double bars provided in the manuscript by Scriabin. Scriabin wrote double bars before b. 92 (coda of exposition), 124 (beginning of development), 206 (recapitulation), 268 (end of recapitulation of second subject. As this last-mentioned place involves a change of time signature, the double bar is technically required, and is reproduced in printed editions, but there is a definite break in the atmosphere here.) The calligraphic beauty and clarity of b. 244–267, a notoriously complex passage, repays study.
Op. 64: Sonata no. 7
The manuscript of Sonata no. 7 is commented upon in detail by Valentina Rubtsova in her notes to the facsimile published by Henle, and these notes are published online. They repay careful study, and Rubtsova gives an account of the other manuscript versions of the Sonata, one of which the present writer has examined in the Glinka Museum. The existence of this text, with its many alterations and differences from the finished version, calls into question the accusation, made by Sabaneyev and since repeated, that Scriabin established a ‘scheme’ of empty numbered bars and proceeded to ‘fill’ it with music. While numbers were clearly important to the composer in establishing a ‘crystalline’ form, the procedure of composition was far more complex than that, as the painstaking work shown in these manuscripts reveals.
Ill.5 is a reproduction of the first page of Scriabin’s earlier draft, with the remarka ‘Prophétique’ for the opening ‘fanfare’ motive. This marking, later rejected, gives a sense of the gesture of this musical idea, which is essential to the close connection of the Sonata with Scriabin’s idea of the ‘Mystery’, something he discussed with Sabaneyev. While visiting an exhibition in London’s Tate Gallery of paintings by the English artist George Frederic Watts (1817–1904), the present writer was struck by the convulsive, ‘prophetic’ gesture depicted in Watts’ ‘Jonah’ (1894), a painting which is reproduced online. The performance of these opening bars needs to be as striking and dramatic as Watts’ painting.
Op. 63, 2 Poèmes
In the second of these short works, some l. h. notes in the chords in b. 6 and 7 have been erased; these notes are relocated to the upper stave, where they belong musically, and marked m.g. (The m.d. in bar 7 is a characteristic slip, rightly questioned by the editor). The top note of these chords is shown in the manuscript as f natural and was so published in the RMI edition. Zhilyayev, who had discussed this passage with the composer, corrected this to f sharp. The first notation shows how essential the gesture of hand-crossing was to Scriabin’s conception of the sonority here. Some pianists make the simultaneity of sounding of notes into a priority, but a letter by Scriabin to Belaieff which has been dated to December 1894 shows that spreading of chords was essential to his conception at times (such spreading was, in any case, far more prevalent at that period than now). In this letter, Scriabin writes that the ‘wide chords’ in bb. 9-10 of the Impromptu op. 10 no. 2 ‘must be played by the left hand alone, for the character of their sonority in performance depends on this.’
The Scriabin facsimiles which have been made available in Russia and America are invaluable sources of information and inspiration, and studying them brings us just a little nearer to the composer. It is hoped that the notes above will encourage players and music lovers to investigate them, and also that more facsimiles may follow in the future.
Simon Nicholls, 2016.
 This edition was the basis of those of the sonatas, preludes and etudes reprinted by Dover, though some of the editions chosen for reprinting contained errors not present in the complete edition. Dover did not reproduce the essential information that nuances and rubatos given in brackets in these editions, notably in the op. 8 etudes, were from instructions given by Skryabin to Mariya Nemenova-Lunts while she was studying with the composer.
 This edition was republished in limited numbers by the Scriabin Society of the U.S.A.
 The letter is dated ‘19th April’ by Scriabin and dated to 1896 on the website. The edition by Kashperov of Scriabin’s letters (A. Scriabin, Pis’ma, Muzyka, Moscow, 1965/2003, attributes it to 1897 (p. 168–169, letter 144.)
 This letter is given by Kashperov (op.cit.) on p. 492–3, letter no. 545.
 Scriabin: Sonata no. 5, op. 53. Urtext and facsimile. Muzyka, Moscow, 2008.
 Boris Pasternak, An Essay in Autobiography, trans. Manya Harari, Collins and Harvill, London, 1959, p. 44.
 I am grateful to the distinguished scholar of Russian literature Avril Pyman for pointing this out (private communication). The French text was added by hand by the composer to the proofs of the first edition (information from the notes by Christoph Flamm to Skrjabin: Sämtliche Klaviersonaten II, Bärenreiter, 2009, p. 43), but perhaps we should trust Scriabin’s Russian, his native tongue, rather than his French in this case.
 Ibid., p. 44.
 Muzgiz, vol. 3, commentary, p. 295. Christoph Flamm, loc. cit. The printed version supplied in the Muzyka edition of the facsimile adds the ties in dotted lines, following Muzgiz. It is certainly tempting to make the ‘correction’: most pianists play the tied version, which persists in many editions. But such bringing into line of parallel passages should not be done automatically.
 Valentina Rubtsova, notes to facsimile of Scriabin Sonata no. 5, p.57.
 Cf. n. 1, above.
 Cf. Paul Badura-Skoda, Interpreting Mozart on the Keyboard, trans. Leo Black, Barrie & Rockliff, London, 1962, p. 290 for a brief discussion of one example of this problem. Brahms wrote against any attempt to improve on Chopin’s orthography at the time of the preparation of a new complete edition of the Chopin piano works (letter to Ernst Rudorff, late October or early November 1877, quoted in Franz Zagiba, Chopin und Wien, Bauer, Vienna, 1951, p.130.) All this comment is made about a single accidental because the orthography of Scriabin’s late music is such a wide-reaching, fascinating and important topic, perhaps seen by some students of the music only as an irritating difficulty of reading, and this is one small example of it. For a discussion of Scriabin’s orthography and its significance see George Perle, ‘Scriabin’s Self-Analyses’, Music Analysis, Vol. 3 no. 2 (1984), p. 101–122.
 Skrjabin, Klavierwerke III, ed. Günter Philipp, Peters, Leipzig 1967.
 Ibid. Philipp notes the variant in an editor’s report, p. 98.
 Christoph Flamm discusses Scriabin’s remarki, and comments that the composer accepted with indifference the publishers’ treatment of his upper or lower-case letters (op. cit., p. 42). Nonetheless, these small ms. differences can be infinitely valuable suggestions to the performer. Flamm points out that even the size of the letters in which a remarka is written can be of significance for the performer.
 Scriabin spoke good French, but accents sometimes go missing in his writing. This circumstance could perhaps be compared with his tendency to miss out accidentals.
 The present writer has read a gramophone record review in which this famous bass note was described as a ‘wrong note.’
 The Austrian piano firm Bösendorfer added a few bass notes to the range of its largest instruments. Apart from making Ravel’s bass notes possible to ‘correct’, the bass strings add to the resonance of the piano. No such advantage attaches to an addition to the top of the keyboard.
 Noted by Darren Leaper.
 Cf. n. 15, above.
 Muzgiz, vol. 3, commentary, p. 296.
 Kashperov, op. cit., p. 87.