The short version
This post started out as a bit of excitement at finding that the source for the Gorsky Fille mal gardée pas de deux male solo was Czibulka’s Scène de ballet Op. 268 (1882), and that it was available, together with the preceding adagio, online at the national library of Spain. (To get the full pdf from that site, click on the green “download” icon third from the left at the top of the page).
Within 24 hours, though, I had found the whole pas de deux (piano music) available on IMSLP, where all of it is wrongly (I think) attributed to Hertel. I say “I think” because who the thing is by is a long, and so far never-ending story.
The long version: how I found it
The strangest coincidence happened on Friday: on my way out to work in the afternoon, I was idly flicking through a pile of old ballet scores that someone had given me a long time back. The first one I looked at properly (because it had no title page, so I couldn’t see what it was) declared itself to be a pas de deux from La fille mal gardée, with “Cziboulka” at the top, and a copyright notice by William McDermott on the bottom. It didn’t look like anything from Fille that I recognized, until I turned a couple of pages, and found a waltz that seemed familiar. Now where had I heard that before? The galop at the end also seemed familiar. It was time to go, so I put it down and left.
An hour later, I had an email from a colleague asking if I could identify and source the music for a variation, as in the video here (with a Youtube link). I clicked on the link: not only was it exactly the music I had just been looking at, but the video was also the one where I (now I remembered) had first heard the music. Had I not picked up the score that afternoon, I would have had no idea what the music might be.
I wouldn’t have known what the music was from the video either, which says it’s Hérold. I hunted around IMSLP for Czibulka scores, but couldn’t find anything that resembled this music. Then on Wikipedia I saw a list of works by him that included Scène de ballet, and wondered whether that might be it. With an opus number (268) the journey to a result speeded up, via an old recording on archive.org, and finally to a digitized copy of the piano score at the Biblioteca Nacional de España.
Czibulka Op. 268: Scène de ballet (1882), the source for the Gorsky “Fille mal gardée” pas de deux
This isn’t the source for Lise’s variation that follows (see 7:05 in the clip below). One forum poster suggests that it’s a supplementary variation by Drigo for The Harlem Tulip, and cross-referencing this with the Marius Petipa Society’s page on Harlequinade pas de deux, which contains the same music as an interpolation, and cross-referencing that with a video on YouTube, confirms this as true. That leaves the coda as possibly from Armsheimer’s Halte de cavalerie.
The Petipa Society’s page on Fille is wonderfully informative, but they attribute the adagio to Hertel and the male variation to Drigo (from La Fôret Enchantée). Either they’re talking about a different solo, or it’s not correct. I have a sneaking suspicion however that Czibulka might have just put his name to the Hertel adagio: it looks and sounds suspiciously not like Czibulka—too early, florid, and harmonically sparse. But is that really likely? Hertel (1817-1899) and Czibulka (1842-1894) were contemporaries, and the Scène de ballet was published in Berlin where Hertel’s ballet was composed. Another possibility—since Hertel’s work was based in part on Hérold’s (1791-1833) earlier score— is that perhaps both Hertel and Czibulka borrowed the adagio from Hérold. But I can’t find a trace of it in the score for the Hérold Fille at IMSLP, and it still doesn’t explain the waltz.
So much for the classics. With all the power of Google and the WWW and libraries and public funded companies and so on, these musical details of what one imagines to be the pillars of the repertoire are still more opaque than if the ballets had been done by monks in the 11th century. It’s a classic case of what Pouillaude in Unworking choreography refers to as part of the “ontological weakness” of dance—the fact that a ballet like Swan Lake (his example) or Fille mal gardée (mine here) seems to crumble and disappear into different “productions,” and cannot be pinned down as a “work” in any stable sense. It’s one of the most fundamental truths about ballet classics, yet seems to be part of no-one’s curriculum, at least not from a musical perspective.
Retracing my steps
In the previous post, I wondered just how much detail is too much when you’re documenting sources and how you found them. With things like this, I think it’s valuable to retrace your steps. With all my favourite finds (#1 being the Giraud Franz solo), there’s an element of sheer luck, guesswork, and then the grind of leafing through page after page of score in the hope of finding gold. Of course, I always check Matthew Naughtin first of all (but he doesn’t mention Czibulka). Wikipedia and YouTube, and The Petipa Society are invaluable. Spelling variations and opus numbers unlock doors. IMSLP is the most lovable musical resource ever, but sometimes you have to manually check archive.org and the catalogues at big libraries to find stuff that hasn’t been uploaded there. If you’re lucky, you can take search for images, and see the front cover of a piece of music, and a link to a library site. And lastly, there’s Shazam, which helped me to identify the recording that appears so often in these videos: it’s Ballet Gala, original Bolshoi Theatre Orchestra conducted by Georgi Zhemchushin. Don’t look there for musical references, though—the composer is given as Hérold.
After nearly 24 hours on and off trying to identify and locate the missing music (the non-Czibulka pieces) I did what I should have done earlier, which is check under Hertel at IMSLP, where, under the German title of Das schlecht bewachte Mädchen, the entire pas de deux was waiting all the time. No mention of Czibulka, Drigo, or Armsheimer though, even though it’s a Soviet publisher who must have known better, surely?
The most curious part of this for me is that the score I had with “Cziboulka” written on is dated 1959, which means that the information has been around for at least 60 years. If you search for Czibulka and “Fille mal gardée” you will find an entry on Suzanne Knosp’s inventory of ballet music publications at the University of Arizona. It’s for William McDermott’s book of Five pas de deux (1985), including this one, with the music correctly attributed to Czibulka, Drigo and Hertel. There’s a lesson to be learned there, though I’m not quite sure how to formulate it, but it just goes to show, there’s a lot to be said for keeping a static HTML list of your ballet treasures.