In the middle of the previous post on music in Karsavina’s memoirs Theatre Street, I found myself diverted by a question that has bugged me and several others: when did the piano, rather than the violin, become the instrument of choice for ballet classes? No sooner had I written the post than I found an enlightening article by Galina Bezuglaya, author of probably the two most comprehensive books on playing for ballet . I have only just ordered the 2017 book, but it seems like the article is probably one of the chapters from that. I’ve moved the section on violins and pianos from the Karsavina post to here and expanded it with new things from Bezuglaya’s article .
The transition from violin to piano
What started this digression was that a few weeks ago, I read in Carol Lee’s Ballet in Western Culture that in 1910, the French had been “amazed” that Cecchetti conducted his classes for the Ballet Russe with a pianist, which Lee said had become “standard practice” in Russia by this time:
In Russia the violin de poche, which had provided the musical beat for dance practice since the eighteenth century, was set aside in 1883 when one of Petipa’s assistants, Alexei Bogdanov, was transferred to Moscow to head the Bolshoi Ballet. In an effort to upgrade the troupe, Bogdanov introduced a number of reforms, one of which was that all classes should have piano music just as all rehearsals were conducted to the accompaniment of the piano.
Lee says that this practice was later introduced for classes at the Imperial companies, and after that, the schools. Her source for this claim, which I have never anywhere else, is a conversation with Erkki Tan who had trained at GITIS in Moscow in the 1950s. It’s an interesting story, but it sounds odd in the context of two other claims about pianos and violins in the ballet class. First, in her conference paper on the subject of the transition from violin to piano at the Sound Moves conference in Roehampton in 2005, Kyoko Murakami quotes Joan Lawson as saying that it was Gorsky who abolished the fiddle from the school in Moscow when he took it over in 1915 . Second, John Lanchbery claims that the first ballet to be rehearsed to piano in Paris was Messager’s The Two Pigeons in 1886 . So on the one hand, why would the French have been surprised to see Cecchetti using an instrument that had been adopted in rehearsals almost a quarter of a century earlier in Paris, and on the other, why did Gorsky need to abolish a practice in Moscow (violins for class) in 1915 that had been the subject of reform there in 1883?
I begun to wonder whether the violin for dancing classes was something that hung on longer in the theatres and vocational schools than it did in every day dancing schools, just as now, it’s the vocational schools and companies that still use piano while most of the world is using recorded music. By the end of the 19th century, the world was awash with pianos and pianists, all needing to be put to some use . Whether they were really preferable for the task of teaching ballet, who knows. As Ronnie Hynd wrote, back in 1961, in a wonderful article about dancing and music, if it’s Tchaikovsky and his predecessors, you barely need a pianist at all: “A great deal of role teaching can be done to a sung approximation of the music” . Perhaps a detail from Nikolai Legat’s memoir, documenting the moment when he began to take over Christian Johannson’s classes provides a clue:
He used me as his assistant and delegated work to me long before he retired. From him I acquired the very essence of his school, and found myself imitating many of his manners, even to the preference for accompanying my instruction on the violin, rather than the piano.
The preface to the book records Legat as playing tunes for class on the piano or violin, which made me think that perhaps the use of one or the other elsewhere in the schools depended on the teacher, their playing ability, preference and habit. Reading about the reverence that dancers had towards Johannson, it’s hard to imagine someone from Facilities turning up, taking his violin away, pointing to a new Steinway in the corner and saying “sorry mate, it’s all piano from now on.” Bronislava Nijinska’s memoirs give another small clue: describing events of 1907, she describes the relationship between her brother Vaslav, and former fellow-student Anatole Bourman. They had ceased to be friends once they were no longer in the same class at school.
Recently, however, he had taken up with him [Bourman] again, and had asked him to play the piano for the dancing lessons he gave. But this was only occasionally when either of his regular accompanists, Leni Gontcharov on the piano or Kolya Issaev on the violin was not available.
I’m assuming she means that it’s Nijinsky who used Bourman as an accompanist for his classes, not the other way round, but whichever it is, clearly as late as 1907, a class could be accompanied by either violin or piano. Nijinska also reports (p. 211) that Kolya Issaev was the violinist at rehearsals for the pas de deux from Swan Lake with Anna Pavlova in the same year. I wonder if that rehearsal was with violin alone, or violin and piano? Or two violins with a repétiteur score?
These tiny glimpses made me realize that I had been asking the wrong question (“when did ballet teachers stop using violins?”). Changes like that do not happen overnight, and in a sense, they aren’t so much changes of instrument as changes in behaviour, and you probably need to look at individual teachers and their practices rather than institutional policy for an answer.
Some time after I wrote this post, I discovered an article about a very early version of “music while you work” at the Rowntree chocolate factory in 1905, which opens with this bit of historical detail:
In May 1905 a contribution to the in-house magazine of the Rowntree company dealt with the introduction of half an hour’s singing for young women covering chocolates at the factory: ‘morning visitors to the Works have occasionally been somewhat astonished to meet a violin in the cream-room corridor about 11 o’clock, looking very frivilous and out of place at that prosaic and work-a-day hour. Accompanied by the violin, women workers would sing hymns as they worked.
Perhaps it bears no relevance to ballet classes in Russia or Paris, but I feel it’s significant that it was a violin, rather than a piano, that accompanied the singing. It says to me that it was still not out of place in 1905 to lead a musical event like community singing with a violin. Given the popularity of the piano at the same time, it lends some likelihood to the idea that there was a period when both violinists and pianists (or teachers who played either or both) around the turn of the century might have accompanied classes.
Galina Bezuglaya to the rescue
The most useful pointer in all of this is Bezuglaya’s article, published in the Bulletin of the Vaganova Academy (see previous post for more on this great journal). The main part of the article is a discussion of published piano improvisations for class: in the pedagogical works of Friedrich Zorn and Nikolai Legat who supply the accompaniment themselves, Sofia Brodskaya’s music for Vaganova’s Fundamentals of Classical Dance, and Maria Pal’tseva’s music in Lyubov Yarmolovich’s Classical dance. See Nina Revskaya’s online library of these texts for more. Incidentally, Pal’tseva can be seen in the film The Children of Theatre Street (see earlier post with a link to the relevant clip).
At the beginning of the article, Bezuglaya briefly discusses the transition from violin to piano in the schools of the Imperial theatres in St Petersburg and Moscow. Until the end of the 19th century, she says, most ballet teachers would usually accompany themselves on the violin—the notion of an “accompanist” is something that comes later. But here’s the vital detail: the administration of the Imperial theatre schools made a budgetary provision for accompanists in all ballet classes, but since the teachers had the right to supplement their own salary with the music budget, many of them were more than happy to do without a musician [cited in ]. So you can imagine that if you could get by accompanying yourself on the violin (like Christian Johannson, Platon Karsavin, or Cecchetti, for example), you would probably do so.
It’s interesting that Bezuglaya, even with the resources that must be available to her, finds it hard to pin down a precise date when the changeover happened, but based on the published reminiscences of dancers and ballet masters of the time in question, she thinks it probably happened gradually, somewhere near the end of the 19th century, beginning of the 20th. Valentin Presnyakov for example remembered having violin accompanists for his classes with N. I. Volkov in 1888, but that they were soon replaced by pianists. In the Moscow school, one of Gorsky’s innovations was to replace the violin with the piano for class some time in the 1900s, but there is no sign of the date Joan Lawson gave of 1915.
If teachers continued to use the violin even after the piano had been brought in to replace it, this was not simply to save money on musicians, Bezuglaya thinks. Teachers of ballroom dance were happy to pay pianists, because there was no shortage of printed repertoire for them to play—polkas, contredanses, galops and so on. The trouble with ballet training—and remember this is the school of the Imperial theatre we’re talking about—is that it was a private, closed-off world that pianists knew nothing about, so it was not a simple matter of pulling in pianists from the street and sitting them down to play. What was needed for ballet classes was not a “repertoire” of composed dances, but an ability to improvise, in however basic a musical style, for the exercise had just been set, taking on its temporal structure, rhythm, dynamics, mood and articulation. Possibly, Bezuglaya says, the violin-playing dancing masters had been able to improvise simple tunes adequate to their teaching aims, whereas that skill had been lost through lack of use by pianists who’d spent too much time trying to become virtuosi, to the extent they couldn’t even harmonize a melody at the keyboard.
Friedrich Zorn’s Grammar of the Art of Dancing
As an example of the crossover between violin and piano towards the end of the 19th century, Bezuglaya draws attention to the difference between editions of Zorn’s Grammar of the Art of Dancing: in 1887, the text is illustrated with single-stave, treble clef violin tunes, either composed by Zorn, or borrowed from works by Weber, Bellini, Meyerbeer and Auber among others . The 1905 American edition, however, has a supplementary booklet of the same tunes arranged for piano. In all the years I have had this very edition of the book, I had never bothered to look closely at this, or ask why it was there, but as Bezuglaya points out, we might wonder whether this is gives a glimpse of what a ballet master at the end of the 19th century might have played for classes, either on violin or piano. . Bezuglaya uses different examples to mine below. The illustration here is from the piano supplement:
The section signs (§) refer to the relevant parts of the book (this systematic and orderly numbering and cross-referencing method appeals to me greatly, as someone who has had to try and do the same in syllabus books). Here is the relevant page (49) with the violin examples shown:
Vaganova and music(ians)
One of the most interesting insights is that according to some of her pupils, Vaganova was apparently uncompromising in dictating her musical requirements. She wanted absolute correlation between music and movement, nothing less. Let’s just remember here that there is no such thing as “music” and “movement” in a ballet class in the abstract: they are the product of a relation between teacher and pianist, and in the nicest possible way, Bezuglaya suggests that what passed for choreomusical “oneness” and “harmony” in Vaganova’s classes was achieved through a certain subjugation of the music and the pianist. Bezuglaya notes perceptively that in this respect, some 20th century teachers managed to turn pianists into the “perfect instrument” to carry out their creative will, which is not unusual considering the way that dance music generally in preceeding generations had been ordered by the yard, so to speak, by choreographers . As she points out, in a period of enormous change in the dance world generally with regards to relationships between dance and music, ballet teachers held fast to earlier principles. I’m sticking my head above the parapet here, but I think the same is broadly true 100 years later. You could walk into many classes today and think you were musically and culturally in mid-19th century Russia, whatever changes are going on around you.
It’s a bizarre coincidence that just as I had been going through ballet masters’ memoirs (such as Legat’s) scraping off the musical data, so to speak, I find that someone else has been doing the same. I’m going to do a whole post on Legat which will mirror and expand on some of what Bezuglaya says, but while we’re on the subject of music, technology, and change, let’s just take a moment to consider Fokine’s piano rolls.
Fokine and piano rolls
It’s worth mentioning that the turn of the century was also when recordings, pianolas and similar mechanical instruments began to be so commonplace as to find their way in to dance accompaniment, after a long gestation period (the year of the first production of La Bayadère (1877) is also the date when first prototypes of the phonograph and player piano appeared). Mass production of records became possible in 1902, while in 1903, the Aeolian company had a repertoire of 3000 piano rolls, and were producing 200 new titles a month.
Lydia Kyasht, Anna Pavlova and Lydia Lopokova all appeared in adverts for the Angelus player piano in Theatre Magazine 1914. What is remarkable is how our expectations of technology have changed: now we expect fidelity, but in 1914, what the ballerinas were recruited to sell was what Nicholas Seaver calls re-performance : the ability for the user to control nuances of tempo and dynamics, with what is charmingly known as the “phrasing lever.” To some extent, things like varispeed and training apps perform a similar function, but with nothing like the flexibility and nuance that a player piano would have had, without any loss of quality due to sampling issues.
Gladys Crozier notes on page 109 of The Tango and How to Dance it (1913—available here) that it was a rare accomplishment for pianists to play well for dance these days, and is thus enthusiastic about the available gramophone records. But she also recommends hiring a pianola (quite reasonable, at a guinea a month) for similar reasons to those mentioned in the advert in Theatre Magazine:
The music provided by a pianola is excellent to dance the Tango to, a special advantage lying in the fact that the sustaining of the bass notes —so essential when playing a Tango. . . is managed without the last trouble by means of the pianola lever, while owing to the elasticity of the instrument, the player can introduce his or her individuality to any extent into the rendering of any special Tango air.
Teaching in New York in the 1920s and 30s, Fokine charged $5 a class, or $20 for a half hour private lesson (that’s about $75/$300 in today’s money), but he didn’t spend much of that on musicians:
A self-taught musician, in class he would usually sing as accompaniment. Sometimes he would put rolls on the player piano which sat in the alcove of the ballroom and on occasion he would sit at the piano and play a phrase-always a phrase, never a note.
While we’re at it—Fokine and music is a subject that needs a post all of its own, but let’s just say that to describe him as a “self-taught musician” is nonsense. His memoirs contain dozens of passages which testify to his musical training and literacy. Here he is describing life at the Imperial ballet school:
We were given music lessons twice a week and were at liberty to choose the piano or the violin. I took both. I have always been a fanatic about music. . . I found time to play the balalaika and other instruments for musical circles and in an orchestra, and to orchestrate and to copy music, and even to conduct.
I wonder what Fokine had in his piano roll repertoire in New York of the 20s and 30s? (For those who have never seen a player piano before, see the clip below). Such a shame that these details get lost and forgotten because they seemed unimportant at the time (I guess you’d be wanting to get the most out of that class, given how expensive they were, not wondering what tune was playing).
Improvisations becoming works
Bezuglaya’s work opens up a whole other world of research on music in ballet practice, and the Bulletin of the Vaganova Academy is an extraordinary resource that has no parallel in the UK. Once you start following the footnotes, you find dozens of articles, books, theses, anthologies and methodological studies that deal with improvising for ballet classes (albeit in Russian). At the same time, there is a perennial problem with providing examples of how to improvise: they end up getting used as works in themselves. This happened, Bezuglaya says, to the music by Maria Pal’tseva included in Yarmolovich’s book of 1968, which provided three sample classes, intended to exemplify for pianists the principles of accompanying classes so that they could go and do likewise. Instead, they have resounded for more than three decades in ballet classes as new pianists use them as material.
Looking at Zorn’s book, the variety of possible accompaniments around at the turn of the century, and reading about what was probably quite minimal accompaniment in the case of Johannson and others, I wonder if the problem isn’t that we’ve become rather too obsessed with the idea that music in class ought to be good quality music in itself when it’s improvised; rather too fixated on particular sounds or versions of things. But that’s a topic for another day!