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

On publishers’ mistakes in the Rite of Spring

The Jerusalem Symphony conductor Frederic Chaslin keeps finding errors in Stravinsky’s published score. So which edition should he use?



rite of spring score rite mss

Dear Norman

I’m writing you because, like many conductors, I will be celebrating on May 29th the 100th anniversary of the “Rite of Spring”.

I have been long complaining and fighting with publishers about the many mistakes that one can find, especially in French opera that are beyond description. Still, in a score like the Rite of Spring, knowing how precise Stravinsky was for details, it is unlikely you could find any single mistake. If you do, and you are Pierre Boulez and have a TV crew nearby, you’d be famous for eternity (remember the movie by Rolf Liebermann). I don’t have a TV crew but found a couple of things that Stravinsky probably dropped here and there in his various compositions, but those are minor details.

Now, my problem, and the reason of my message, is the orchestra parts. The new set that Boosey  & Hawkes is sending (and for which they are charging a lot) is just not possibly the same that was used in the past. The mistakes that one would find in the parts are not only A LOT, they are really bad.

An example? You open the 2nd violins, and right on their first entrance, Stravinsky repeated their figure 3 times. He wrote that figure in a crescendo and for the next times, he wrote “sim”, which the publisher should understand and “simile”, SIMILAR. Instead, they wrote DIM, diminuendo. So just one letter different, and the musicians will play the opposite of what they should. And it goes on and on until the end with small details that matter.

rite mss2



Another example? the violas have, around the figure 184, a lot of grace notes. In the original score, those notes are at the beginning of the bar, were all accidentals are cancelled. In the part, they are inside the previous bar, being impacted by whatever accidental was inside that bar. That makes a difference! Again for the viola this is a small example among many.

So, I can assure the Jerusalem audience that they will have a “Kosher” version clean of mistakes. I strongly advise my colleagues who are using a new material to review it, and to the publisher, I suggest to discount their rental.

Best wishes

Related Posts Plugin for WordPress, Blogger...


  1. This situation reminds me Paavo Berglund on 2001. He was supposed to conduct the São Paulo Symphony (OSESP) on Shostakovich S. 8°. However as soon as He got the score, many mistakes started to appear. There wasn’t time enough for change and He decided to rehearse and play Brahms S. 4°, that scores were available. He declared that the publisher was totally guilty (I cannot remember the name).

    • John McLeod says:

      Again – Boosey and Hawkes! Also I, and probably hundreds of others are still waiting for a facsimile score of ‘Le Sacre’ which they offered at a special price (£79) early this year. 14 weeks ago the money was taken from my account, only to be told last week (after much enquiring) that the score would not now be ready until the end of June. So they must be sitting on thousands of pounds acquired since March!

      • rvslkal says:

        Actually, the standard western editions of Shostakovich are from Sikorski. Not sure who represents them in the Western Hemisphere, right now, but it used to be Schirmer. The actual plates were (I think) taken over directly from VAAP.

  2. Jonathan says:

    In these days of easy computer setting its astonishing that awful hire materials can remain an issue. But the onus is on someone providing a service or product to provide something of merchantable quality and I am not sure that all conductors / orchestra managers are aware of the power of complaining: I once made a publisher send three different sets for a “standard” 20th century work. Only the third was of anything close to the quality one might reasonably expect but needless to say we discovered some pretty astonishing mistakes in rehearsal – even though the set was not new by any means. In short, for works in constant use for over 50 years, there is simply no excuse for providing poor materials so if its not good enough, demand replacement or a refund.

  3. Shengyun says:

    Slonimsky wrote in Perfect Pitch that Prokofiev found 53 errors on first page of a proof of Stravinsky’s work…he didn’t check proof at all!

  4. Tom Gossard says:

    That’s because Stravinsky preferred to get paid to getting the published scores and parts right. Given the sorry state of music publishing at the time, and still as in this case, is it surprising there were errors? There always are in the publishing business no matter what is being published, it’s part of the territory and a necessary part of the process, like it or not. Publishers in Stravinsky’s day always held the upper hand and many if not most composers were relatively poor to penniless in their vocations.

  5. We entered the Firebird (complete ballet) into the computer and produced parts from that. It is time-consuming, but you will probably end up with a cleaner and more accurate score. The score and parts (which you can see on IMSLP) are full of mistakes. We also have an as yet not proofread Rite score and in copying we found some mistakes,although it sounds like most are in the parts. In any case, modern parts produced from computer generated scores have the virtue of having the score and parts matching (or at least they should). It saves many hours in rehearsal and is always worth the time and trouble.

    • Jonathan says:

      This is highly practical and producing and circulating decent editions of many works would probably be of more real value to the original composers than the persistent use of awful materials. Unfortunately, though, in the case of Stravinsky, in most parts of the world producing one’s own materials would constitute a breach of copyright – is there anywhere it does not? In any case, importing and using them anywhere where copyright is respected could be problematic.

    • Andrew Condon says:

      The Eulenburg score of the Firebird Ballet (1910) published in 1996 ( a new Urtext edition according to the publishers) still perpetuates many errors; the brand new computer set parts based on this edition still have errors, and so far as I am aware, the promised Critical Report to this edition ( that could have addressed the many inconsistencies and anomalies that remain in the score) has never materialised .

  6. Publishers’ lack of interest in the quality of scores and parts is not new – even for their biggest moneymakers. Stravinsky’s work has been a particular problem from the start. For many years, the parts for Les Noces were particularly notorious. The problem with The Rite of Spring is more complex though – and not just a matter of wrong notes – although that’s bad enough.

    Like Petrouchka, from nearly the start of its career, Stravinsky kept making small changes to the Rite of Spring score – that is until the 1940s when he began to rescore the piece. He only completed the reorchestration of Danse Sacrale and that was published separately by Associated Music (In the US). The question is, which Rite of Spring does one perform: The score as it was originally performed in 1913, the score as it was first published after WW I, the score with the various changes over the years after that but were incorporated into by Boosey onto the original RMV plates (with LOTS of misprints and mistakes), the early 1960s re-engraved edition that Boosey published without consulting Stravinsky, or the combination of one of those with the 1940s rescoring of Danse Sacrale? Regarding the 1960s edition – one has to ask on what basis they selected from the earlier changes? And, the problem with the 1940s re-orchestration is that there is a very noticable change in the style of orchestration from the rest of the earlier versions of the piece.

    Boulez and Michael Tilson Thomas, I believe, have used and recorded the 1913 version, Bernstein used to perform one of the earlier versions with the 1940s rescored Danse Sacrale. Stravinsky did the same. If anyone doubts the number of changes in the score over the year (and this doesn’t include misprints), all one has to do is to compare almost any two recordings at random. One has grace notes here and there, others don’t. There are lots of other distinctions as well. I would imagine that most people perform it from the re-engraved version without any investigation into the various versions and changes.

    The real scandal here is that at 100, there is no scholarly edition of this keystone of 20th century music that checks the score against all of the manuscripts and published editions, taking into account of all of the changes made and/or reinstated. My suspicion is that conductors will never be in agreement as to which changes they want to incorporate, but they should be able to make an intelligent choice. These things matter. At least, they should know that there has never been any “final” version of the piece.

  7. Oh for heaven’s sake you pompous conductors. Stop whining. We orchestra players deal with misprints on a regular basis. We usually just fix them without bothering you. Most of the time conductors don’t even notice
    the minor errors. Why waste good rehearsal time on this when unless it’s a really difficult one to solve the individual players can just research it and fix it.

    Of course there are errors in Stravinsky. This isn’t rocket science. Anyone who’s played in an orch, for any period of time knows this. The conductor being written about here makes it sound like he’s had some divine
    revelation. He is reinventing the wheel. Rite has been played plenty of times, and most of the errors have been noted by the players, even if they haven’t been officially corrected by the publisher.

    You want to know about errata, consult Clinton Nieweg, Principal Librarian Emeritus of Phila. Orch. and head of MOLA (Major Orch. Librarians’ Assn.). Clint has compiled a generation’s worth of errata lists for every piece imaginable. Out of curiosity someone should ask him to share what he has on Rite. I’d bet it’s a bible.

    • Chumalovsky says:

      Bravo ,Carmen ,well said. This could cause a right riot…………..of spring,of course.

  8. From my own side of the fence, I have always felt that a composer’s re-edit of a previous work is the final say. That being said, should we actually perform the “original” version of anything, be it “Le Sacre” or any of the Bruckner symphonies? I don’t see people rushing out to present the first version of Schumann’s d-minor, despite Brahms’s insistence that it was better than the composer’s “rewrite” (or are they, and I’m just in a cultural backwater?)

    I have been attempting to work with Boosey editions (of both wind and orchestral works) for my entire career and continually find them rife with errors. They are not the only culprits, but seem to be among the most egregious. As one writer noted, in this day of easy computer transfer of score and parts, one would expect error-free materials, especially when paying top dollar for rental and/or purchase.

    • I don’t think there is much to prove one way or the other that a composer’s second thoughts are always an improvement over his first. Composers make changes for all sorts of reasons but there is always the risk that later-in-life changes will carry with them jarring changes in esthetics as well. The orchestration of the 1911 Petrouchka is opulent and difficult to execute. But is also clearly an esthetic product of its time. The 1947 re-orchestration is leaner, clearer, easier to execute, and certainly more sophisticated in places, but it is just as clearly a product of post WWII Stravinskian esthetics. Which is better? It depends on your point of view.

      Rachmaninoff, whose style of composition in the 1930s became lean and not unaffected by neoclassicism, was constantly at odds with the esthetic of his earlier works. He made drastic cuts to the 3rd and 4th Piano Concertos, as well as the 2nd Symphony, but hardly anyone accepts these anymore. Why? If one weighs benefits of concision against being able to hear the complete melodic developments, the latter wins.

      You correctly note that few may perform the amazing, original version of Schumann’s 4th Symphony, but this is probably due to the fact that few people even know it or have seen a score. And let’s face it, most people prefer what they are familiar with. But this is also the exact same reason that people prefer the first version of Mendelssohn’s Italian over its re-composition. In both cases, though, I think the initial versions are far superior. And, regarding the Schumann symphony, don’t be so quick to knock it. Not for nothing did Brahms put his relationship with Clara Schumann at risk by publishing it.

  9. It is amazing how many misprints there are that performers have been playing and recording for years. I’m thinking of a particularly glaring wrong melody note in Villa-Lobos Etude No. 5 that has been played for decades. I did a whole fair-sized post on misprints, including a really bad one in the Chopin B flat minor sonata. Here:

  10. I don’t know if any of you are members of MOLA (Major Orchestra Librarians Association) but we have spent a lot of time creating and putting on line errata for numerous publications and editions (only accessible to MOLA members). I don’t know if the new Boosey Rite has been done yet (I’m retired and no longer active in MOLA activities) but this is a tool of inestimable value for correcting scores and parts. You can make use of work already done by scores of librarians around the world. I highly recommend that the Jerusalem Orchestra join MOLA and use this wonderful tool. Patrick Zwick, Utah Symphony Librarian & MOLA Webmaster (retired)

  11. Roberto Gonzalez says:

    Best thing is to buy a set of the original parts, NOT BOOSEY, which was done by Stravinsky to get back the copyright, but not for any real musical reasons, and work from those. The Conductor’s Guild of the ASOL has a set of corrections to the original score and set of parts. The other reason Stravinsky made some changes was because he could not conduct the original, and so rebarred some places, according to the story I heard. Luck’s Music Library has a set of the original print for sale…

  12. My recommendation would be to use the Boosey 1967 re-engraved edition and correct any parts you have to it. While it might be interesting to hear the heavier scoring of some passages of the first RMV editions, Stravinsky was apparently unsatisfied with the original orchestration. Besides, the heavier original scoring would sound clotted in the reverb of typical large concert-halls, unlike the drier theatrical sound of the original performance venue. Copies of the very first RMV edition of 1921 are hard to come by as it has not been reprinted by any of the normal sources (Kalmus, Dover, IMSLP etc.). As far as I know, Harvard has not put online its copy of that score, something I’ve urged them to do.

    I also recommend a close listen to both of Stravinsky’s Columbia recordings of the piece. Both the 1940 version, not to mention the stereo version utilizing the 1943 re-scoring of the Danse sacrale, have EXTREMELY important changes in the chord articulations of the Danse. For example, the 8th note in the 3rd bar after [142] as notated in the original scoring is performed as short as the 16th note in the previous bar in both recordings. The 1943 rescoring has both notes notated identically (as 8th notes, the fundamental rhythmic unit of the rescoring). This tightened articulation is apparently something Stravinsky REALLY wanted. He also wanted a SLOW tempo for the start of Part II, which is commonly performed far too fast.

    There are a couple of touches from the original scoring of the Danse sacrale that I wish IS had preserved through all versions. First is a guero scrape into the last chord of the piece (Bernstein keeps this). Second are the alternating pizz and arco string chords in the final section (reduced to all arco in the later revisions). While it may look like the pizz chords would be inaudible considering what is else is going on at the time, they actually cut through the texture and the irregular alterations in coloration the create add another layer of rhythmic complexity (Dutoit’s Montreal recording is one of several that keep the pizz chords).

    ALL conductors should view Millicent Hodson’s reconstruction of the original Nijinsky choreography . It is eye- and ear-opening. To begin with, It explains several puzzling aspects of the score (such as the otherwise strange molto allargando at 3 bars after [116]. And it serves to confirm the general rightness of Stravinsky’s metronome markings. The tempi chosen should be danceable in the original choreography. The Nijinsky choreography can be found in several incarnations on YouTube. Avoid the Mariinsky performance which is too prettily 19th-century balletic. The Joffrey 1989 performance is better at conveying the score’s modernistic primitivism.


    • Michael Schaffer says:

      The reconstructed choreography is certainly interesting to watch and maybe even eye-opening – but how “ear-opening” is it really? While they made great efforts to reconstruct the original choreography, sets, and costumes, they didn’t pay much attention to the score, to reconstructing what it may have sounded like in 1913.

      That is the subject of a highly interesting article written by UCLA musicologist Robert Fink: “The Rite of Spring and the Forging of a Modernist Performing Style” which tries to trace how the score and the way it was performed changed over the decades.

      • If one insists on “what the score sounded like in 1913″ you would have to include aspects of original-instrument performance practice. Among other things, at that time the strings were probably still all gut and played without constant vibrato. French bassoons through at least the first half of the 20th century had a very distinctive, almost saxophone-like timbre. And there’s the vibrato of the old French and Russian schools of french-horn playing (now largely disappeared). The shape of the Theatre des Champs-Elysees precludes it from having the long reverb times of a traditional shoebox-shape concert hall — the acoustic was therefore probably rather “dry” but clear. There are places in the score where any or all of these factors would far outweigh the actual pitches being played in the sonic impression created. Besides, a pit band playing a brand-new piece of the Rite’s difficulty and complexity in the face of audience disturbances probably produced a performance below the standard of the many youth and college orchestras that now tackle the Rite.

        I may also point out that while Stravinsky seems to have changed the orchestration on a regular basis, his metronome markings have been retained through all the printed editions and perhaps even from the autograph original scores (now nearing facsimile publication from Boosey). They are thus integral to the work and major deviations from them are musically equivalent to performing wrong notes.

        Boosey Rite facsimiles:
        Boosey brochure with images from Rite autographs containing metronome markings:

        Those interested in the history of the score and the choreography should immediately wing themselves to Paris to attend the 100th anniversary events at the Theatre des Champs-Elysees. Wish I could be there:

        A virtual tour of the stage house with a view of the hall itself is here:

        The Theatre’s website says that the May 29th performance of the Rite (Gergiev, Mariinsky, Nijinsky choreography) will be carried live by Arte. Let’s hope it also appears online.

        • Don Drewecki says:

          Sixtus writes: “If one insists on “what the score sounded like in 1913″ you would have to include aspects of original-instrument performance practice. Among other things, at that time the strings were probably still all gut and played without constant vibrato. French bassoons through at least the first half of the 20th century had a very distinctive, almost saxophone-like timbre. ”

          Yes, and you hear that in the wind playing of the first Stravinsky recording, from 1929 or ’29, in Paris. (Reissued on Pearl 25 years ago, it should also have been included in this new Sony Masterworks boxed set as it was also issued by Columbia Records in the U.S..)

  13. There are many many errors in the original Russe de Musique orchestral score of The Rite of Spring in several printings, plus revisions. For years the most readily available score was the Kalmus reprint; there was a new Soviet engraving that was reprinted by Dover; and then there’s the 1967 revised and re-engraved Boosey & Hawkes study score, which in no way should be regarded as a last word. There are textual problems with all of these scores. Complicating the issue is the question of the piano-four-hands score, on which Stravinsky worked at the same time as the full orchestral score, 1912-1913; there are significant differences there as well, some with no ready explanation. It is only to be expected that Stravinsky continued to make revisions in the score almost to the end of his life. Robert Craft lectured on these revisions but a complete recension and correlation of all the sources is an ongoing project not soon to be completed. What we’re waiting for now is the facsimile edition of Stravinsky’s autograph scores, promised for later this summer.

    I posted the above ¶ to Facebook this morning and respond now to this thread here. I don’t have my library where I am at the moment and can’t point to specific chapter and verse, but I cite one example of Rite of Spring complexities: in the Prelude, on the fifth page I think, there’s a new motive announced by the oboe, with an ornate twittering counterpoint with a solo alto flute. This alto flute lick is ten notes in the orchestra score, and twelve notes in the piano-four-hands score.

    The Firebird Ballet score that I first studied was the Broude Bros. reprint of the original Jurgenson edition, which was riddled with misprints. The 1964 Soviet score, re-engraved, cleared up many errors. As for Les Noces, the original Chester edition was one of the worst I ever saw of any score. The parts were if anything even wrose. I published about 200 corrections to the Les Noces score in the Conductors Guild quarterly journal. The new edition, also by Chester, is excellent.

    Some years ago (you can find it on my website) I published a meticulous review of Denis Herlin’s meticulous edition of Debussy’s Nocturnes. I lauded Herlin’s edition, mentioning that no other major orchestral work of the past century had offered so many editorial problems during its publication history. I now wish to withdraw that statement; the score of The Rite of Spring is surely more complex and difficult in that regard.

  14. Boosey is one of the worst publishers I know for inaccuracy over orchestral parts. I remember giving the first performance of Andrzej Panufnik’s Piano Concerto in its 1986 revised version and at the first rehearsal we were stopping quite often to sort out mistakes in the newly-copied parts – major errors such as the percussion parts missing at the start of the slow movement.

    I also remember constant disruption when playing the piano during a rehearsal of Britten’s War Requiem in 1995 – again, it was newly-copied parts by Boosey, with countless mistakes; precious rehearsal time was being wasted simply because some sloppy copyist hadn’t been able to keep his or her mind on the job in hand. Boosey’s hire charges are far too high for such second-rate performing material.

    The increase of copyright protection from 50 to 70 years was depressing news to anyone patiently waiting for scholarly editions of some of the 20th-century scores which Boosey has partially misrepresented for decades.

  15. With so many discrepancies- not merely publisher errors but also changes by the composer- there is something called cy près. The fact that Boulez uses the 1913 score gets you back to the original performance, while other changes provide different interpretative values. Some details are more critical than others, but its also easy to get lost and not tell the forest from the trees. In the end it boils down to interpretative judgment, and I like the idea of a handcrafted score after careful study.

  16. Rosalind says:

    It would be great if Norman could get a response from Boosey and Hawkes about this. Given the cost of hiring music it is quite astonishing that noone apparently proof-read the materials…

    • Hear hear, Rosalind. I would like to tell Boosey that their edition of Copland 3 really sucks. They’ve revised the score but not the parts so there are tons of errors.This wastes rehearsal time, practice time for musicians and is a disgrace considering how much they’re charging for the parts.

      One good feature of Boosey I’d like to point out is that they now feature online perusal scores for their major works. This is quite a blessing. You register, log in & can have a look thru their online scores.

      • Oh, so, you’re the guy (or girl) who, under the disguise of “Carmen”, says: “Oh for heaven’s sake you pompous conductors. Stop whining”…. And now you’re whining as well about Copland… Being a pompous player, may be?
        Well, I’m a composer in the first place and it’s as a composer that I posted this message. And I can guarantee you that, NO, the players don’t “know what they have to do”. When I sent them the list of about 230 mistakes I found (important ones, notes, rythms, nuances) they were quite happy to put them in. Because it’s an orchestra that didn’t play that piece a lot. But when I assited Boulez 20 years ago and that he corrected important articulations mistakes on orchestras who thougt they played it right (and started arguing loud until Boulez played them 2 Stravinsky’s own recording showing he was right) , they too were very happy to understand what’s true and what’s wrong. The thing is, there are the players who only care about their pay check, and the others, who care about their mission. Call them pompous if you want, “Carmen”….

  17. Robert Fitzpatrick says:

    I do not have the expertise of the learned folks commenting above,,,,but, that won’t prevent me from making an observation. Some people consider the Kalmus edition (omg yes, Kalmus!! imagine dat) of Le Sacre from 2000 edited by former Philadelphia Orchestra Clint Nieweg (member of MOLA) as the latest, most authentic edition. Please comment of this heresy. Grazie mille, Spahseebah bolshoi.

    • Robert Fitzpatrick says:

      Clint was the Phila Orch Librarian who succeeded the legendary Jesse Taynton.

    • Robert,

      Clinton Nieweg spent a lifetime collecting errata from not only Phila. Orch. musicians, but from libriarians of every other US orch. as well. These are the mistakes noted by the musicians when they play a piece, reported to their orch. librarian.

      Clint has been using these errata lists for revised editions of Kalmus works. So actually, it’s very possible that this 2000 revised Kalmus edition is the most accurate.

      I personally use a Kalmus edition which must be pre-2000 because it doesn’t bear Nieweg’s name. It’s not particularly bad. Just did Rite 3 wks. ago & I don’t remember any errors in the part. If this has been revised even more by Nieweg it’s probably a very good edition.

      Rite is public domain now, I believe, so many orchs. own their own sets of parts. Corrections can be made permanently to these parts.

      Boosey is another story. The problem on rental works (Boosey is all rental) is that orchs. just have them for a limited time so you fix the mistakes for your own performance and send the parts back. The the parts go out again to another orchestra who does the same. Sometimes the corrections remain in the part from rental to rental but in theory all pencil markings are to be erased before returning the parts, so many corrections are lost.

  18. Rob van der Hilst says:

    This is in fact an old Problem, solved by the composer himself. Compare the printed orchestral score – with its many mistakes – with Stravinsky’s own, very accurate 2pianos-reduction et voilá. ‘Translate’ all these right notes into the not-right printed parts of the orchestral score and your Rite-paradise has been reached. C’est si simple comme bonjours. Good luck!

  19. For the record, stuff published pre 1923 iis public domain in the US, so the first publication of the Rite qualifies. There are complex idiotic and boring rules about using your when playing copyrighted material which are frequently ignored epecially by school groups.

an ArtsJournal blog