an blog | AJBlog Central | Contact me | Advertise | Follow me:

Tchaikovsky’s wrong note? Latest: It wasn’t Tchaikovsky’s….

My pal Stephen Hough is eating supersized portions of humble pie at the neighbourhood health café this morning.

Yesterday, you may recall, he discovered a blue-pencilled correction in an early Tchaikovsky score of the piano concerto which altered an odd-sounding solo F in the flutes to a B-flat, far more convincing.



Today, Stephen accepts that the hand with the blue pencil is not the composer’s. It probably belonged to Hans von Bülow (above, with Bismarck), the exceptionally irascible pianist (also husband-in-law to Richard Wagner), who gave the concerto’s first performance and established its lasting popularity.

But hang on…. is that really humble pie that Stephen’s eating across the road? It looks to my practised eye more like low-cal black-forest gateau. He’s not repentant at all.

What’s this he writes?  So the direct evidence is weaker – this manuscript (especially the blue pencil mark) is not a primary source in Tchaikovsky’s hand; but the circumstantial evidence seems to me stronger: it is a copyist’s mistake, probably corrected by the eminent performer, and more likely than not from the composer’s own instructions.

Ah, so the performer knows best…..

Read Stephen’s pie-recipe here. I’m going for my slice now.

hough lebrecht

Related Posts Plugin for WordPress, Blogger...


  1. Michael Hurshell says:

    “Husband-in-law” – no. He typically referred to himself as “Cosima’s widow.” :)

  2. Theodore McGuiver says:

    We do this all the time in opera. HvB’s one-time champion turned wife-stealer is a particularly fertile pasture for this kind of discovery.

  3. I think Stephen’s carefully expressed arguments (in favour of the proposition that the F natural is an error) are quite cogent and persuasive (both in yesterday’s and today)’s version. Obviously it is difficult to know for sure what the composer intended but that applies to an awful lot of issues.

    • I left some comments on Stephen’s blog about this, which may be of interest here. The natural expressive climax of the opening phrase of the flute solo is the B double flat, which is played three times in bars 11-12 before resolving to an A flat and leading to the piano’s entry. The expressivity of this B double flat is emphasised by its being the highest note in the flute’s melody, and this is perhaps why Tchaikovsky altered the third note of the flute line to an F (instead of the B flat which occurs elsewhere in the movement): if he had used a B flat as the third note of the flute melody, it would have undermined the expressivity of that B double flat, as that climactic note would no longer be the highest note of the phrase, having being pre-empted by the B flat. Using an F avoids this.

      The reason Tchaikovsky can use the B flat when the piano enters (quite apart from its exquisite effect in ‘decorating’ the flute’s statement) is because the piano’s version of the melody will replace the flute’s B double flat with a rise to D flat in bar 19 – the highest melodic note so far heard in the piece – so the lower B flat will not undermine this expressive peak by pre-empting it.

      Further evidence to suggest that Tchaikovsky did not intend consistency is the coda, where to achieve calm the composer replaces the B flat with an A flat. If we can accept the ‘inconsistency’ of the use of A flat here, why can we not accept the ‘inconsistency’ of the F? And when editors accept ‘inconsistencies’ in Beethoven’s scores as deliberate composition intentions (when I would argue that often they are not) why can we not accept them in Tchaikovsky?

      More evidence supporting the legitimacy of the F is that the beginning of the slow movement represents the first return of the key of D flat major since the opening pages of the first movement, and the flute solo features a subtle motivic link with the opening theme of the concerto, though I suspect that it’s a subconscious allusion, rather than a deliberate quote: in bars 12-13 of the first movement we hear the four-note motiv A flat, E flat, F, A flat played on violins (over a combination of dominant harmony and a tonic pedal point, harmonically similar to the pizzicato accompaniment in the slow movement), with a repetition of the E flat, F, A flat progression following immediately in bars 13-14, recurring again in bars 18-19 and then in 30-31 when the piano enters.

      • John Borstlap says:

        These are the MOST convincing arguments in favor of the F, both in musical and logical terms. It is the expressive context here which determines the ‘inconsitencies’.

      • Excellent … thanks!

      • @Raymond Clarke Excellent point about the B double flat, in as much as the strings don’t go higher than an Ab in the first 12 measures, reinforcing that note as a stable pitch and point of departure. Going one semitone higher has a yearning quality, quite Chopinesque. Having the Bb before that would have spoiled the effect, an anti-climax actually.
        Anyway, it seems that Tschaikovsky wanted the F and not the Bb in the beginning as long as he supervised the 2 piano version that has the F.

  4. 18mebrumaire says:

    Hadn’t realized that Bismarck had studied the oboe da caccia with von Bulow.

  5. How can you be sure it’s an oboe da caccia? It could be a bong in disguise!

  6. The F is not an error. Tchaikovsky wrote it deliberately to contrast the Bb written later in the melody. A question/answer format.

    Remember, Tchaikovsky was a flautist himself and understood that to write a Bb in the initial flute melody would sound too strong, the F makes the melody sound soft. He hardens the melody later by writing the alternative harder sounding Bb.

    In some ways not dissimilar to the ‘wrong’ sounding note Mahler writes in the 2nd bar of the slow movement to his 6th Symphony. The Fb (5th note 1st violin) sounds angular. Had he written a normal F, it would have taken the ‘edge’ away from the melody. We get the F natural in the fourth bar, sweetening the ‘edge’ slightly.

    • Alan’s comment is spot on. Everyone else: bask in the glow of its brilliance.

    • Eduardo Monteiro says:

      Alan, I second your points

    • Andrey Boreyko says:

      Absolutely agree with Alan. Exactly ! Only F natural for the first appearance of this tune

    • Michael J. Stewart says:

      I agree with Alan’s comment too. It’s a contrast which to my ears heightens the piano’s entrance. It makes the ear say ‘hello, what’s this?’ Even though it may not do it on a conscious level. Tchaikovsky’s own 2 piano arrangement bears this out too.

  7. Malcolm James says:

    There are also a few well-known errors which get perpetuated. As an oboist, two which come to mind.

    1. In Bizet Symphony 3rd movement trio. A concert B flat is written in the oboe and clarinet parts, but a B natural is almost always played, which makes the melody more piquant and rustic. Presumably a ‘correction’ by a ‘helpful’ editor.
    2. The the cor anglais solo in Falla’s Three Cornered Hat there is a missing note. Apparently a piano roll of Falla himself playing the pasage on the piano includes this note and the publishers have accepted that this is a copyist’s error and have now published a corrected edition.

    • Lindsay Ryan says:

      Malcolm, we’re currently playing the whole ballet (concert Sat if you’re interested) from Chester score/parts. I’d be really interested to know which note is missing/corrected, as we discovered a couple of errors in the 1st violin parts last night, so I doubt we have a “corrected edition”!

      • Malcolm James says:

        In bar 4 of the solo the second beat should be quintuplets with an A (concert D) inserted at the end, so that the runs always move stepwise and go triplets – semiquavers (16th notes if your American!), quintuplets and lastly sextuplets. Link here.

        As for whether I’m interested in your concert – I’m in Cardiff, Wales. Where are you?

        • Lindsay Ryan says:

          Brockley, London 7.30pm Saturday. A bit of a trek from Cardiff, but you’d be most welcome! Thanks for the clarification – it does appear to be correct in this score and the player’s part. My orchestra is Harmony Sinfonia

          • Malcolm James says:

            Presumably the fact that the work is still under copyright means that the publisher can still control the version that it available.

  8. Martin Roscoe says:

    I’ve asked flautists (and conductors) to correct this on many occasions to B flat. No-one has ever complained (so far)

  9. Hough, in his backpedaling piece and in defense of the B-flat states that “a composer rarely presents a theme in a variant form at the very start, especially a theme which then appears numerous times in unvaried repetition.” But one only has to look at the theme opening the last movement of Tchaikovsky’s Pathetique with its bizarre division between 1st and 2nd violins. Heard with the standard 19th century divided-violin orchestral layout, the descending scale “theme” is actually in neither the 1sts nor the 2nds but is somewhat of an acoustical illusion. And after a verbatim restatement at measure 20, it is not presented in this fashion again through the remainder of the piece. While fascinating, what Tchaikovsky meant by this hocket-like procedure is beyond me.

    • Andrey Boreyko says:

      The instrumentation of this opening theme is very strong proof of necessary to perform Tchaikovsky’s symphonies using divided violins layout. All composers living and working in Sankt- Petersburg at the end of th 19th century knew exactly how the Court Orchestra is sitting. By composing they sometimes used this divided layout for the dialog between first and second violins , like by antiphonal singing of two choirs ( two kliros) in Orthodox Church. Such dialogs we can find also in works of Rimski-Korsakov, Borodin, Cui, etc.
      If by the perfomance of “Patetique” the first and second violins are sitting together – the pre composed effect of the “birth” of the “urmelodie” from two different lines is unreachable.
      (Sorry for my imperfect English)

      • Absolutely, Andrey. And Mahler. t’s a bad English habit not to split the strings.

        • Split violins is a fetish of mine, as it was for such conductors as Boult. And indeed it is a bad English habit started by Henry Wood in 1911 (he took “credit” for it in his writings). Divided violins continued to be used by most major orchestras through the 1930s when the “modern” seating started taking over in America and in Germany. So the composers for whom divided violins and cellos on the left is the only authentic seating plan include not only Mahler and Tchaikovsky but also to such turn-of-the-century composers as Elgar, Strauss, Dvorak, Bruckner, Rachmaninoff, Debussy, Ravel, Sibelius, Berg, Webern and Schoenberg. Musical meanings for all these composers are altered and sometimes damaged by seating the violins together and having the cellos on the right. There still are no integral CD sets of the symphonies of Bruckner (I haven’t yet heard Blomstedt’s with Leipzig) or of Sibelius with divided violins. As far as I know, there’s never been a commercial stereo recording of the Sibelius 4th done this way. And the Tchaikovsky Serenade only just recently received such treatment. I wrote a article for Opus magazine many years ago on orchestral layout. I’m thinking of rewriting and updating the piece if I could find a place to publish it (and get paid).


          PS: Bravo to Boreyko for his stance. I look forward to hearing divided violins in live performances of his streamed on the web, just as I like finding Halle/Elder and RNO/Pletnev webcasts with the “old” seating.

          • John Borstlap says:

            Excellent comment.

            The reason that 2nd violins have been seated left next to the 1st violins has to do with the more homogenized sound, and also to the sound quality in itself, because if they are seated right, the players have their instruments turned away from the audience which seems to hinder volume and clarity.

            See Norma del Mar’s informative ‘Anatomy of the Orchestra’ , Faber & Faber, 1981 & 1983.

          • Michael Schaffer says:

            Kubelik retained the traditional seating so all his SOBR recordings have split violins and celli/basses on the left. Mravinsky and Temirkanov did, too, so it is still the standard seating for the StPP today. Sinopoli adopted the traditional seating in his recordings with the Philharmonia, e.g. the complete Mahler symphonies, but oddly enough not in his concerts. The traditional seating has made a big comeback in “mainstream” orchestral performance, too, in the wake of the HIP movement. Thielemann always uses it and in some halls, like the Musikverein in Vienna, he also has the basses in one line behind the whole orchestra which provides a very solid and present bass foundation to the sound of the orchestra.

          • Fabio Fabrici says:

            @John Borstlap
            That is an ill guided argument by Norma del Mar. Nothing sounds more homogenized. the Vl2 have less overtones when sitting on the right, but having the Celli playing sideward to the audience instead is a much bigger loss.
            Also the contact between concert master and leader of the cello group is better with the Celli inside, and the communication between concert master and solo cello is the most important one. There a numerous other arguments against seating all the violins in one big section.

          • John Borstlap says:

            These are convincing arguments…. and they seem indeed better than the possible loss of tone quality when they sit next to the first. However, Del Mar did not ADVOCATE sitting them together, merely stated the reasons behind the idea.

          • I agree with all of Fabrici’s points, especially in regard to the cellos — arguably the second most important line in the strings — which are facing precisely the wrong direction when seated to the right. Cellos have f-holes too and radiate much of their sound like violins — perpendicular to their top plates. Seated on the outside to the right the’re aiming their sound toward the 1st violins, not the audience.

            I must also put in a word AGAINST a “compromise” seating that has the violas on the LEFT behind the 1sts and the cellos on the right with the 2nds. This has been used by Gilbert with the NYPO and Rattle with the BPO. I’ve found that both in live performances and recordings of this arrangement that the viola line is LESS audible than when they are seated on the right, behind the 2nd violins.

            Divided violins playing in unison or at the octave interact (via their early-reflection patterns) with shoebox-shaped concert halls in a way that produces an uncanny effect of being enveloped by violin tone, at least when heard from close the center-line of a hall. It is an effect that is physically impossible to achieve by crowding them all on the left and despite the fact that this was supposed to improve violin tone in such passages. It is so striking that I believe Mahler was talking about it when speaking to Natalie Bauer-Lechner: “It must be the effect of some acoustic law that the sound-waves [from 1sts and 2nds playing in unison] encountering each other FROM BOTH SIDES produce such a lively and brilliant tone-quality.” (3 Dec 1899)

            Finally, I must bring up the 3rd dimension: the seating of orchestra on ranks of risers. This was EXTREMELY common throughout the entire 19th century (e.g. Berlioz) and early 20th and can do wonders for the clarity of the winds and inner strings (cellos and violas). Risers are also the only way to make acoustic sense of Mahler’s demands for woodwind players to raise the bells of their instruments (Schalltrichter auf!) which directs the sound over, not into, the heads of the players in front of them. Risers are even built into the stage architecture of two of the greatest halls (the Concertgebouw and Musikverein). They aren’t responsible for those hall’s renowned acoustics, but do help an orchestra interact optimally with the halls’ inherent properties, as does the “old” orchestral layout.

          • andrey boreyko says:

            In Brussels, by Orchestre National de Belgique we are working now about web-streaming. Hope to be able invite you soon to our concerts.
            By the way, using splitting violins i also decided to put violas next to the first violins. In that layout cellos are next to the second (basses behind cellos of course). What i have reached:
            second violins feels themselves much more secure and confident having basses and cellos next to them. From other hand violas are sounding better when they are partly facing the audience, and not only often performing middle voice between melody and bass, but also sitting between first violins (melody) and cellos (bass).

            Last but not least: in my opinion there is no one universal strings layout for all kind of music.
            Certain repertoire, or certain concert halls ,(like Musikverein in Vienna for example) requires or suggest certain sitting order.
            Their is only one criterium: the best possible realisation of composers ideas.

            once more please excuse my imperfect English. I’ll be very thankful if the moderator will fix my possible errors.

          • John Borstlap says:

            That is fascinating. Especially the violas, with their less penetrating power, have then more exposure. Advocates of the conventional platform planning, i.e. 1st and 2nd violins next to each other, and then violas in the middle and cellos and basses at the right, think it is more advantageous to have the entire range from high to low nicely laid-out from left to right. With some repertoire this will give a more homogenous sound.

            But wouldn’t the wide separation between 1st and 2nd violins not seriously hinder ensemble playing? There are so many passages in the repertoire where they play in octaves, I think many more than passages where they operate antiphonal, like Mahler IX. Thinking of the back desks…. not easy, unless the platform is relative narrow, the string group quite small, and the hall not too big.

            And what about the regular imitations of the motive in Schumann IV (1st mvt, I seem to remember) where it is travelling from celli via violas to 2nd and 1st violins? That has one big gesture from right to left but will be broken if the planning is different. In Shostakovich V in the slow movement, the entire violin group is divided not into two, but into three parts, and only for this movement; there, the conventional seating is, it seems to me, the only workable one.

          • The Schumann 4th is a perfect example of how musical gestures can be altered with anachronistic orchestral seating.

            1. The “modern” orchestral seating (all violins on left, cellos on right) was invented in 1911 by Henry Wood. Schumann would have had no idea that an orchestra would ever be arranged without divided violins.

            2. The premier performance of the 1st version of the 4th Sym was with the Leipzig Gewandhaus, which is known at the time to have had divided violins. I seem to recall also that the violin and viola sections stood while playing, and that the violins were reversed compared to later practice (firsts on the right, 2nds on the left). This reversed arrangement was used by Wagner when he led the Dresden Opera and is preserved to this day at Bayreuth.

            3. So passages such as the 1st movment, rehearsal K that Borstlap hears as “one big gesture” from right to left, was performed at its premier (and until the 20th century) as a series of ascending calls and responses (my characterization, there are others possible). Besides, thinking in “big gestures” is more characteristic of Wagnerian and post-Wagnerian practice, again anachronistic when talking of Schumann, or so the Early Music experts who advocate small-gesture “rhetorical” phrasing would tell us.



            PS: The Shostakovich 5th was premiered by Mravinsky and the Leningrad Philharmonic. He always used the old seating plan (split violins, cellos left). As I described above, having violins spread across the width of the stage creates a rich sonic envelopment, which is perhaps precisely what Shostakovich was shooting for in these passages. Similar effects would occur with the divisi string passages in the Tchaikovsky Serenade.

          • John Borstlap says:

            Great to hear all that. The Schumann argument I can entirely accept, it would have been a ‘question / answer’ effect. But the Shostakovich still lumbers-on as a question mark: in this slow movement, he instructs the ENTIRE violin group – so, 1st and 2nd together as one lump – to be divided in THREE. Where would they be seated? Left, right, and in the middle? And would they then be re-seated for the next mvt which has normal division in 1st and 2nd?

        • andrey boreyko says:

          Thank you Norman. Well, of course Mahler too, and many other composers of 18th and 19th century.
          But by performing music of russian composers from St.Petersburg (the end of 19th and beginning of 20th century) conductors should pay a special attention to this issue. Without splitting violins Sheherazade, Spanish capriccio, and many other works of Rimski-Korsakov, Glazunov,Tchaikovsky, are sounding very different, and since we do have stereo recordings– at home also very different.

      • Hope to see you conduct Glazunov 4 one day, Andrey. You’re one of the finest conductors working today and deserve to be at the helm of a great orchestra!

        Remember your Tchaikovsky Suite No 3 at the BBC Proms in 2010, it was a stunning interpretation!!

        Thank you

        • andrey boreyko says:

          Thank you Alan for your very kind words. I am happy if my work is seen and appreciated, because actually it is for me even more important as to be at the helm of the orchestra with Big Name! I love to work even with less famous, or less known orchestras if i feel good chemistry between conductor, and musicians.
          P.S. …and although I never conducted Glazunov 4 so far, you are encouraging me! Maybe one day?!

      • Fabio Fabrici says:

        Absolutely agree with Andrey, not only the Russians but pretty much anything written before 1920 or 1930 had the composer have the standard string layout of 2nd Vl to the right in mind. It can be shown in numerous scores from before Mozart to Strauss how composers expected the orchestra to sit with the violins split left and right in the front.
        The dubious honor to having invented I+II Vl sitting together on the left was the above mentioned Henry Wood and a few years later Stokowski experimented with it in Philadelphia. From there it spread like the plague unfortunately. It made a lot of sense for mono recordings into one single funnel or microphone but not for anything else.

  10. Mindy Kaufman says:

    I always liked and played B flat.

  11. Tchaikovky had other strange habits, in terms of structure, too: the ‘walz’ theme of the beginning of the 1st mvt of the mentioned concerto never returns – if I’m right – and isn’t that extraordinary, in a 19C concerto in romantic grand manner? Even Liszt’s piano concertos are more structurally coherent.

    Maybe we should also take into consideration that 19C orchestras were technically not as flexible as the modern orchestras: slight irregular things which would have given ‘edge’ or would sound a bit ‘out of context’ at the time, can be ironed-out today, like the ‘break’ in the middle of the clarinet range.

    In 18C and 19C for instance composers had to reckon with quite complicated horn slides and their relation to keys; Wagner muddled a lot with valve horns, natural horns and their capacities, but nowadays players handle these things much more smoothly.

    • Fresison says:

      Here are two interesting quotes from Francis Maes’ “A History of Russian Music: From Kamarinskaya to Babi Yar”–

      “[T]he work had another beginning: instead of having chords struck forcefully across the entire keyboard, Chaikovsky gave the melody to the strings, accompanied by arpeggios over a limited range of the piano. This is understandable because the melody is sounded only in mezzo forte; the piano does little more than add color. Where the full chords come from is in fact a mystery. They first appeared in the third edition (1889), and may have been suggested by Siloti. The theatrical pose of the new piano part is most inappropriately treated by many soloists as an acrobatic tour de force. Such an approach often destroys the unique character of that famous page, the particular middle way between grandeur and elegiac lyricism. It would be interesting to restore the original beginning as an alternative in concert practice, if only to discover what Chaikovsky originally had in mind.” (pp. 75-76)

      “The key to the link between the introduction and what follows is once again Chaikovsky’s gift of hiding motivic connections behind what appears to be a flash of melodic inspiration. The opening melody comprises the most important motivic core elements of the entire work, something that is not immediately obvious, owing to its lyrical quality. However, a closer analysis shows that the themes of the three movements are subtly linked. Chaikovsky presents his structural material in a spontaneous, lyrical manner, yet with a high degree of planning and calculation.” [At this point, there's a footnote referring to Jeremy Norris, "The Russian Piano Concerto I", pp. 114-151.] “The use of folk song material is particularly subtle. The first theme of the sonata form of the first movement uses the Ukrainian folk song “Oy, kryatshe, kryatshe . . . .” The middle section of the second movement is based on a French chansonette, “Il faut s’amuser, danser et rire.” A Ukrainian vesnyanka, or greeting to spring, is the basis of the first theme of the finale, while the second theme is motivically derived from the Russian folk song “Podoydu podoydu vo Tsar-Gorod.” A strong motivic link ties all the songs together. Because they are known songs and not newly conceived themes, the relationship between them is often attributed to chance. How-ever, it is likely that Chaikovsky chose these songs precisely because of their motivic affinity, deploying them whenever necessary. The selection of folkloristic material went hand in hand with the planning of a large-scale structure.” (p. 76)

      • John Borstlap says:

        All that is really interesting. Thanks for the trouble to quote these passages. The structural affinities between the various themes must be the reason why that introduction does work perfectly natural, even if it does not come back literally. This is a master composer.

        • @John Borstiap This is probably not a place for involved structural analysis but I assure that the 1st Concerto by Tschaikovsky is not so far from sonata form that Liszt’s concertos were.
          I don’t know what you mean by the “waltz” theme but there is in the development a rhythm in 6, that really doesn’t stand alone as a theme but is derived from previous material.

  12. Surprisingly back in 1958 Leinsdorf with Leonard Pennario and the Los Angeles Phil in a brilliant recording, gives the Bb not the F, to the flute in the beginning of the 2nd movement. There might be a case to adjusting that note to be in line with all the other instruments playing the theme in that way.
    However, the piano actually ends with Ab, F, and repeats the Ab. So it is not unform throughout the movement. I personally like the F at the outset but maybe that’s because I am so used to it…

  13. Glad to see that most of those who are interested in the subject agree with my preference for F over Bb. The verdict seems clear. We have all seen plenty of typos and mistakes made by copyists and occasionally by composers themselves. However, where else have we encountered this kind of error when: 1) a note is altered neither by a minor or major second nor by an octave (an 8va mistake), but by a full fourth, 2) no accidentals – whether wrong ones, superfluous or missing – are involved, 3) there are no miscounted ledger lines here — and all of this with the result that not only sounds good but is truly well justified musically (as shown by several comments above)? Probability of such “error” is infinitesimally microscopic. And on top of all that, the composer’s own two-piano arrangement apparently confirms the correctness of the score, which should be more than enough evidence to end this discussion once and for all – decisively in favor of F.

an ArtsJournal blog