Tag Archives: waltz

A year of ballet playing cards #38 (QC): Prague Waltzes: Soft, strong and very long.

Screen grab of piano score of Prague Waltzes by Dvořák

Prague Waltzes: click the image to download the free piano score

When is a waltz not “a waltz”? Most of the time 

If I ever get to play what I think of as  “a waltz” for class (you know, the rollicking, flowing, swaying kind that has a pendulum swing in it that propels you forward without ever getting tired) , I mentally crack open the champagne. Ninety percent of the time in class, you’re trying to find something that is waltz-like, but not exactly “a waltz.” I suspect the problem is that the waltzes we know from the concert repertoire were made more for ears than legs. I have rarely, if ever, found a suitable moment in a ballet class  to play Léhar’s  Lippen schweigen (“The Merry Widow Waltz”), yet that’s one of the first tunes that comes to mind when someone says “waltz.” Over ten years, many of the posts on this site have hovered around this topic in one way and another, to the extent that I’ve now created a page listing the “waltz problem” posts.  

The sound of three heads turning

Much of the music you’re asked for in class has zen-like conundrums in the specifications. A colleague said he’d been asked by one teacher  for a “melting march.”  Sounds familiar:  I tried to solve a similar problem with what I called a “chameleon-like March by Granados).  Waltzes for multiple pirouettes are similarly taxing: you need something slow, but not too squidgy. Rhythmic, but with space for allowing more turns without sounding naff. Elastic and steppy for balancés, but then with three sharp beats that can signify three “heads” for a triple pirouette.  

If there’s a model for the tune that can accommodate all this, then perhaps it’s the opening theme of  Kaiserwalzer Op. 437 by Johann Strauss II: 


But it doesn’t last long, and it’s played so often for classes, you can only use it sparingly. 

That’s why Dvořák’s Prague Waltzes is such a find. Like the old slogan for Andrex toilet paper, it’s soft, strong, and very, very long. If you’ll forgive the comparison, the design problems of pirouette music and toilet paper are not so dissimilar. Beats in waltz music need a softness combined with a tensile strength such that they can hold together and stretch without breaking, but also separate with a quick tug when you need them to.  And here you have it: pages and pages of pirouette music that does all the right things (though I’ve made a few minor cuts to make it class-ready).

Prague Waltzes is a useful model of what “waltz” can mean. This composition is evidence that waltzes don’t just go “1 2 3 1 2 3” — there’s a whole world of varied accents and tempos and rhythms within a single phrase. Most significantly, in my view, there’s a lift/accent/length/weight, call it what you will, in the middle of the bar rather than beginning, and often a sense of direction towards the third beat, not the first; sometimes there are three separate gestural beats in a bar, not three subsumed into one. Prague Waltzes also provides  many examples of how to vary and extend a waltz idea when you’re improvising. 

I also love the title, having spent every 14 of the last 15 years playing for the International Ballet Masterclasses in Prague. It’s a city I will associate forever with ballet, playing for some of the best and nicest people in the ballet world. I wish i’d had this music for some of them, however, considering how many hours of pirouettes I must have played for. 

Tempo for Prague Waltzes

I left the allegro vivace  on this arrangement out of deference to the orchestral score, but to me this doesn’t sound right given what’s on the page, and so the metronome marking range is mine.   I like the tempo that Jirí Belohlavek takes it with the  Prague Symphony Orchestra (I also like to think they must know what they’re doing with this Czech music).   For class, you could take it even slower, and pull it about in different ways as necessary. Belohlavek plays around with the tempo quite a lot for the sake of concert interest, but the opening sections are the kind of tempo which works well for a lot of pirouette exercises. 


The mysterious case of the Lyrical Waltz


I’ve just had an email from a teacher, asking me what I understand by the term “Lyrical Waltz.” Short answer, I don’t understand anything by it, but the long answer is that I’m rather fascinated by how a term like this can gain such currency over a long time, without apparently having much meaning. 

Lyrical waltz: a potted personal history

The first time I heard the term “lyrical waltz” was when I started work at the RAD back in 1986. I think it was something that teachers had been told was a meaningful musical term to use to pianists. I used to improvise waltzes that started with  a dotted quarter note + three eighth-note pattern (as in the Sleeping Beauty lilac fairy attendants example below). I soon ran out of ideas. I think the reason I associated this pattern with “lyrical” was because somewhere in a syllabus book there was an exercise that had “lyrical waltz” as a tempo indication, and that’s roughly how the music went. 

Screengrab of the piano score of Tchaikovsky's Sleeping Beauty, Lilac Fairy Attendants

Lilac fairy attendants from Sleeping Beauty (Tchaikovsky). When I hear “lyrical waltz” I think of this rhythmic pattern of dotted quarter note + three eighth notes. But I’m not convinced that’s good enough.

What—if anything—is a lyrical waltz? 

Over the years, I have tried to work out what, if anything, a “lyrical waltz” is in musical terms, but have only come up with more questions. 

  • Does it mean something that has the quality of a song? That doesn’t really work, because there are plenty of songs that have a bombastic quality.
  • Does it have a melody that is song-like, rather than being motif-based like the Act 1 waltz in Swan Lake, or the opening of waltz of the flowers in Nutcracker, where you can hear the composer at work, rather than the singer. However, as soon as you start singing these tunes, they have a song-like quality, they’re singable. Back to square one.
  • Does it mean something that has more eighth-note motion than 1-in-a-bar feel? Not an infallible criterion, because there are 1-in-a-bar waltzes which could be described as lyrical, and eighth-note ones which aren’t.
  • Does it just mean slow? I don’t think so, because teachers who have ever asked for this didn’t (I think) want something ponderous
  • Does it mean something where the melody takes precedence over the accompaniment, i.e. something like La plus que lente by Debussy? Up to a point, but if teachers use  the word “waltz” at all, I presume they’re expecting more rhythmic predictability than this.

Lyrical waltz—a pedagogical category only?

By “pedagogical category” I mean a term that has arisen from a teaching context, but has little relation to the world outside, but has somehow stuck. Whoever started using it may have had a particular waltz in mind, like the “Lyrical Waltz” of Shostakovich, from which they extrapolated a category, without giving it much thought. I think this happens a lot—where people like a single tune, not realising that what they like about it is particular, not generic. Take La cumparsita which people have sometimes used as a generic template for “tango” — when it’s about the only tango that goes like that, and in fact, was never a tango in the first place, but a march. As an illustration of this in practice, a colleague told me of a class where the teacher had sung a tune while she marked the exercise, and then said “But don’t play that. Play something similar.” You guessed it: after a few try-outs, she said “You know what, just play what I sang.” 

Incidentally, this is the opposite of that odd, ballet-only scenario where a teacher will ask for “The same thing” by which they don’t mean literally the same thing, but something that is in metre, tempo, style and feel the same, without being, you know, the same. This is where the everyday German distinction between das Gleiche and dasselbe is useful.  There might be an interesting intersection here between musicology and everyday ballet class practice. In Music, Imagination and Culture (1991), Nicholas Cook writes of the tendency to “hear works as individuals rather than as exemplars of a type” (p. 147) and that this is  a “defining principle of the aesthetic attitude,” citing Dahlhaus’s Analysis and Value Judgement (1983, pp. 13-14). In my experience, ballet pianists are much more attuned to attuned to what dance forms are as a genre than classically trained musicians. Ask the latter for “a polonaise” and they’ll play an exemplar, of which they probably only know a couple of the Chopin compositions, without being aware of the things that make it a polonaise in the first place. 

Lyrical waltz—or little waltz?

One teacher I play for often asks for “A little waltz” and for some reason, I know exactly what she means, though it could also be the tone of voice and gesture that conveys the idea. “Little” to me here suggests something in moderate tempo, moderate volume, not bombastic, not grand, with a smooth melody line, perhaps like the Tchaikovsky E flat major waltz Op. 39 , or the Little Waltz by Teresa Carreño.  A piano piece, rather than an orchestral number reduced for piano. A miniature. Little is a more productive and meaningful term for me than lyrical, though I’m still not convinced it helps. I’m also referring mentally to particular pieces that have an overall quality elicited in performance more than composition. 

Lyrical—just a name, rather than a category?

I searched around for “lyrical waltz” on Google, and then for Valse Lyrique. Once you exclude Shostakovich or Sibelius, it’s not a huge list, so the idea that there was once a whole category of waltzes called “lyrical” is suspect (though you’ll find quite a few of them on ballet pianists’ albums, which supports my theory that it’s a pedagogical term, not a real-life one). 

In the US Library of Congress Catalog of Copyright Entries 1945 (Music) New Series Vol 40 Pt 3 No 10 there are more compositions in the index with the word “Valse” in the title than “waltz,” and only a handful with the term “lyric.” When you look at the list of adjectives associated with “valse,” (see below) apart from lyrique including erotic, beige, parfumée, you begin to wonder whether any of them have much meaning, except as a way of flogging a generic composition as if it might be particular. Perhaps lyrical is doing the work of organic, natural, new, advanced, healthy, free-from! in food-labelling. If we’re fooled by food labels, I’m sure we can be taken in by sheet music covers.

Picture of the index from the US Catalog of Copyright Entries for Music 1945, showing a list of compositions including the term "valse"

Extract from the Catalog of Copyright Entries (Music) 1945, compositions with the title “Valse”

Postscript: Is “a lyrical waltz” something to do with the body, not music? 

Once I’d written this, I began to wonder whether the term “lyrical” has some purchase with dance teachers because of the genre of lyrical dance, in which case maybe it means “the kind of music I can do emotionally charged slow bendy dance to.” That opens the field up more, without the need to get too metrical-technical about the music. 


A year of ballet playing cards #22: A schmaltzy waltz by Kéler (9h)




Click to download the score

Elastic or expressive timing is a feature of music in ballet classes that I’m rather nostalgic for. When I meet teachers who still do it, I feel as if I’m back home again.  Recently, I’ve found that I need a particular category of waltz for company classes with multiple pirouettes, where the tempo needs to be forgiving, accommodating, like stretch fabric jeans that can be a couple of sizes larger than the label suggests. Schmaltzy is what we would have called it a few decades ago, but it’s a long time since I heard that word.

How I learned to love the schmaltzy waltz again

When I first started working in ballet, I’d be slightly appalled if a teacher wanted part of an exercise slower than the beginning, to make space for a different kind of step or more pirouettes. Now I enjoy the interaction with the teacher, enjoy the challenge to get the tempo change right and still make it sound musical.

You can’t do this with any old waltz, and particularly not with the kind that has an overwhelming metrical level, rather than switching between several through the course of the tune. I’ve put Deutsches Gemüthsleben in the “hearts” section because it has places where it’s in six (see various posts on triple meter), but the opening is very definitely in three. Sections E and F do what the Swan Lake Act 1 waltz does, which is to drive the accent to the second bar and fourth bars. At the ends of sections, the three-in-a-bar is suddenly more marked. In other words, what appears to be simply “3/4” has several different kinds of metre both internally (within a single phrase) and from section to section.

I’d never heard of Kéler until I read that he was the source of the Hungarian dance that made Brahms famous when he plagiarized it in its entirety, believing it to be a folk tune (like folk tunes just precipitate in the ether like ectoplasm, huh?), but I’m glad I found him. See more in Nancy Handrigan’s wonderful thesis On the Hungarian in Brahmsespecially pages 55-56, and for the proof, listen to the original Kéler below (scroll to 2:56″ if it doesn’t start there automatically).

I haven’t found a performance of this Deutsches Gemüthsleben Walzer that I like, or that does justice to its potential for elastic timing (some of which Kéler writes into the score), but it’s remarkably similar to the waltz in Der Rosenkavalier, which is the feel I’m looking for (see below – scroll to 5:40 if the technology doesn’t make it start there automatically).

It might also not escape your notice that both Kéler’s melody and Richard Strauss’s have a lot in common with The Lonely Goatherd from The Sound of Music, and also — i noticed much longer aftewards — with the “Waltz of the Flowers” from The Nutcracker.   Perhaps these falling thirds against an unchanging major floor is music’s answer to Gemüt, and perhaps Gemütlichkeit is the best way to describe character of the schmaltzy waltz that you need for a certain kind of pirouette: music that’s warm, roomy, comfortable, supple, supportive, like an armchair by the fire in an old pub.

A year of ballet playing cards #20: A luscious big waltz (Talisman coda)

Piano score for the Talisman pas de deux coda (waltz)

Click to download the score

This is probably the nicest and most useful waltz for grand allegro I’ve found in a long time. It just sounds like ballet. You can use it straight off the shelf, and it’ll work instantly, and I love it.

The same goes for the adage from the Talisman pas de deux (my last entry) which I tried out in class today. It sounds just like all the things teachers seem to have in their heads when they mark adages, yet so few pieces actually deliver.

It also has within it a brilliant example of the difference between “normal” waltz metre and truly triple metre. The first and last sections are in “normal” waltz metre, i.e. in what we could otherwise notate as 6/8, with a weaker second main beat of the bar, whereas the middle section is truly triple, with accents every three. It’s hard to think of a better example to make the point with.

It’s not the cleanest score I’ve produced, but I’m trying out my little Akai LPK25 for the first time, and getting used to using laptop commands (i.e. without a numerical keypad)  for a big editing job in Sibelius. It’s hard work, but I’m so glad to have finally done what I’ve been meaning to for years, and buy a little touring keyboard for inputting scores. I remember reading once that Czerny had so many projects on the go that he’d have a room full of desks with a project on each, and go round each one for an hour each, and then move on to the next one. It hasn’t got to that yet, but I found myself rather naturally using one side of the table for PhD work, and the other side for playing work. It makes it so much easier to put things down when I get in.

Image of laptop and mini MIDI keyboard

My desk in the apartment in Prague where I input the Talisman pas de deux coda


A year of ballet playing cards #16 : Esmeralda male variation music (3h)

Esmeralda male variation music (clip of the piano score)

Click to download the score (pdf)

You can never have enough grand allegro, and this is handy because it’s in a class of pieces that are ballet music, which means that you have to be careful where you play them, but on the other hand, it’s repertoire that’s not often performed, so either people won’t know where it’s from, or they’ll smile and go “Isn’t that…??” and you look good because you know weird stuff that you found on Youtube. The solo is at 48’46” in the clip below. It should start there automatically when you click, but if it doesn’t, drag the slider to the correct time.

See also: 

Esmeralda male variation music: the meter (for geeks only)

This should really be in the Clubs suit, not Hearts, because it’s actually a truly triple meter, not the dodgy six-eight kind—the phrases end on the eighth count, not the seventh. What fooled me was the melodic phrasing, which is in two bar units, which definitely feels duple.  But look more closely, and not only are the cadences on 8, but also the harmony changes every bar, which strengthens the case for truly triple metre even more. Also, the introductory vamp before the first jump is one bar long, not two, which aligns somewhat with what William Rothstein has to say about “Franco-Italian hypermeter.” I transcribed this from the recording, so I don’t know whether in fact Drigo did write in 6/8, in which case the single count  vamp would align with that theory even more.  If it were the case, then the “extra” bar in the middle is not extra at all, because the melody begins on the half-bar in a 6/8 (but don’t try actually playing it that way in class).

On the other hand, it could just be a kind of compositional economy: given that you’ve already got an eight-bar phrase of entrance music, you don’t want to prolong the vamp any more than absolutely necessary, so keep it short, if you must have one.  Maybe it’s there  to provide the dancer with a run-up into the first jump (the vamp-like nature of the music telling the audience that what’s happening isn’t yet dance, just preamble to be ignored.

Once you start thinking about Rothstein’s theory (see other posts here and here) it makes something apparently as unimportant as an introduction suddenly fascinating, and it opens up all kinds of possible discussions about metre, grouping, phrasing, accent, and so on. For me, dance makes those questions particularly obvious because you’re dealing with accents and trajectories that happen in time, but they aren’t “musical” in the sense of being tied to time signature or accent. It’s like seeing a landscape compared to an ordnance survey map.



A year of playing cards #33: A deathly slow waltz (7c)

A very slow waltz for ballet class: Death of Nikiya from La Bayadère

Click to download score

Thanks to Grant Kennedy in Australia for this as an idea for adage/ronds de jambe, anything turgid, anywhere that you need a very slow waltz for ballet class. As it’s from the ballet La Bayadère, you’d want to avoid it for company class if it’s currently in the rep – but as it’s a principal solo, I reckon you could get away with it as long as it’s not a recent memory (especially if it’s in a men’s class).

The threeness of the very slow waltz for ballet class

Over the last few years, I’ve become increasingly convinced that ballet classes require just about every kind of triple metre  under the sun except what most of us know as a waltz. Polonaise, mazurka, polka mazurka, kujawiak, sarabande, chaconne, redowa, to name those I can remember. But even things that look like waltzes on the surface in ballet often have non-standard features: slow tempo, 8th note rather than quarter note motion, and here’s an odd one: a lean towards the second bar of each two-bar unit, not weight on the first. For the prime example of that, think of the famous Act 1 waltz from Swan Lake – it’s all about the first beat of the second bar, and there’s nothing at all on the first beat of the first bar. I can think of several other examples in the ballet repertoire. (For more on my obsession with triple metre, see earlier post).

I reckon that this waltz from La Bayadère is marginal to the waltz repertoire by virtue of  its extremely slow tempo. There are, it’s true, several valses lentes in the concert repertoire, but La plus que lente by Debussy is only just a waltz, and not really that slow. The nearest relative of the waltz in today’s “playing card” would be Sibelius’s Valse Triste. But even that has livelier moments. Nikiya’s death waltz is deathly slow, and every darn beat in the bar has weight.  This is a bar where you’ll wait forever to get served.

And if you’d like to see what they do in the ballet.

A year of ballet playing cards #30: A useful Polka mazurka

Segment of the polka mazurka for ballet class

Click to download the free piano score of Ziehrer’s polka mazurka

“Such serviceable flowers” The polka mazurka as the chrysanthemum of the ballet class

You’ll probably look at this and think it’s a bit dull. I can live with that, because I know you’re going to thank me one day when you get some really awkward exercise that needs to be like doing a hill-start at traffic lights every bar. As Miss Brodie says of chrysanthemums in The Prime of Miss Jean Brodie “Ah chrysanthemums. Such serviceable flowers.” A withering remark, but nonetheless, if serviceable is what you need, then a polka mazurka (or a bunch of chrysanthemums) is the thing.

Truly triple meter and the polka mazurka

I’m very glad that I made the distinction from the start in this series between 3s that were really more like 6/8s (hearts) and 3s that are really 3s, because this is a perfect example of truly triple meter, and why polka mazurkas, twee and 19th century as they are, are often much better for  pirouette exercises than waltzes. See previous posts on triple metre, and waltzes and mazurkas.

I’m making up all the articulations in this transcription, because I have no score to refer to, so I’m just writing down what I can hear. Listen to the recording, because everything about this piece is about articulation and tempo, rather than notes (and that’s why it looks a bit dull on the page).

Wurf-bouquet means a “throwing bouquet” (presumably like the one you throw at weddings), and my guess is that there’s some musical gesturing going on in the melody, where the three-note motif (and the way the anacrusis is held back in performance) is meant to represent the preparation and throw of a bouquet. It’s the kind of thing we get used to doing in music all the time if you play for class. I haven’t transcribed the ending of the piece – it goes back to the beginning, and then spends about a minute winding down, which is not much use to anyone in class.

There are some interesting similarities between this and the opening of the Grand pas classique in Paquita, which will tell you pretty much for certain that what we have in the ballet is a polka mazurka. That being the case, as lovely as the production below is, I’m guessing it’s probably a bit racier tempo-wise than the original might have been. I could be wrong – one of the useful things Zorn told us was that Russians waltzed faster than most of the rest of Europe at the end of the 19th century, so maybe that applies to the polka mazurka too. The long-short-short-long pattern in some of the bars is identical, and is worth keeping in your tool-box as a means of slowing down waltzes for pirouette exercises.

I’m a week behind, as I was at the 2nd Music and Consciousness conference in Oxford last week. Given that other people manage to con prisons into releasing them by setting up email accounts using an illicit mobile phone, you’d think I ought to be able to blog my way through a conference, but I needed a midi keyboard and more time and brains than I had at the end of the day. I’ll catch up soon.


A year of ballet playing cards #29: A Chaconne and a grand dance (3c)

Image of the chaconne by Purcell

Click to download the score of Purcell’s chaconne

Why is the chaconne so rare in ballet class culture?

I’ve no idea. Given that ballet’s history is supposed to have its roots in Italy and France, it’s odd that things like minuets, chaconnes, bourrées, sarabandes, gavottes and so on are some of the rarest things ever to be heard in a ballet class.  If there’s any cultural hegemony going on, it’s Austro-Hungarian: the waltz and the polka.  Of all the music that isn’t played in classes, it’s the chaconne that seems to miss out most. I don’t think there’s a reason why, except that the only kind of three people can think of in a hurry is the waltz and the mazurka, and those rhythms get embedded in the technique, I suppose. Can there really be something about ballet that requires the waltz? Or is it that chaconne requires learning a new rhythmic trick, and ballet habits are remarkably resilient?

Install a new chaconne instead of your old waltz

I discovered a while back that if you’re lucky, and the exercise isn’t too slow, you can replace the thick, porridge-like stir of a waltz played too slow for ronds de jambe with a Chaconne like this (albeit played rather too slow for a chaconne). It has a push on 1, but not the feeling of a wellington boot sinking into mud that the slow waltz has. The dotted rhythms and true triple metre (see earlier posts on the topic of triple meter) keep the music moving.

There’s something hypnotic yet interesting about the variations on the ground bass: it’s amazing how  much harmonic interest Purcell squeezes out of a single 8-bar bass line.  It’s also handy that there’s loads of it – 16 eight-bar phrases. I’ve put rehearsal marks on each one so you can pick and choose depending on the length of the exercise, but beware of – it’s the only time the phrase ends in a dominant.  There are some great recordings available, but I chose the one below because it has some baroque dance in it as well – which helps to give an idea of what a central tempo for this could be, even though it bears playing more slowly.

By chance, I started to input this just before going to a wonderful concert of the Shostakovich 1st violin concerto at the Barbican, and I was thinking of how the first of Shostakovich’s 24 preludes and fugues is a little chaconne, rather like the one in Purcell’s The Fairy Queen. Then In the violin concerto there’s also a kind of chaconne-like passacaglia in 3, which I felt peculiarly prepared for, as if inputting the Purcell had been a kind of mental warm-up.

On wikipedia there’s a wonderful list of other compositions that have used the chaconne as a basis. Well worth looking at for other ideas, if you liked this one.


The Lanner in the Stravinsky

The Steyrische Tänze Op. 165 by Lanner - source of one of the ballerina tunes in Petrushka

The Steyrische Tänze Op. 165 by Lanner – source of one of the ballerina tunes in Petrushka

It’s well known that Stravinsky borrowed tunes from Lanner for the ballerina in Petrushka. I knew about the Schönbrunner waltz Op. 200 [link to score], but I was pleased to find the other tune on page two of the Steyrische Tänze Op. 165 [link to score].   This is why I love IMSLP – it gives you a chance to recover sources like this. I can remember reading a book or programme note about  Petrushka that said in a supercilious tone Stravinsky was caricaturing the facile melodies of Lanner’s waltzes, and in a sentence like that, Lanner gets pushed further to the bottom of the heap of composers that one is not supposed to like, or even look at – you can just rely on some secondary source to tell you what to think. I’m not saying that when you look at it again, you realise that Lanner was a master composer and this is a wonderful piece – but I do like to have the opportunity to hear this music without the modernist composer’s, musicologist’s or critic’s sneer all over it (hear it for yourself here.) It opens with a waltz in 3 bar phrases (like the Glinka Valse Fantaisie), and is really rather nice.

The only reason I found it was because I’ve realised that the Steyrische (or “Styrian”) has got just the right degree of turgidness for a certain kind of ronds de jambe exercise, so I was hunting around IMSLP to see if there were any good ones to add to the repertoire. I’ve played the Schönbrunner waltz several times in class, and no-one’s ever said “That’s Petrushka!” so I wonder if they’d notice if you played this. Of course, if they did, the nice thing is that they’d think you were playing Stravinsky, so you could put on your Stravinsky face and make out that it was really hard.

I’ve also played “Po dikim stepyam Zabaikalya” which is also quoted in Petrushka, and no-one’s ever noticed either the folk song or thought it was from Petrushka (they’ve not noticed the other folk tune – Я вечор млада во пиру была –  that I play sometimes, either).


A year of ballet playing cards #15: A rumpty-tumpty waltz (Oh! Oh! Antonio)

Oh! Oh! Antonio - first line of the music

Click to download the score of Oh! Oh! Antonio

I like it when teachers set grands battements on what I call a rumpty-tumpty 3.  My favourite pieces for this kind of exercise is the Zarah Leander song Davon geht die Welt nicht unterand Hands, knees and boomps-a-daisybut all good things must come to a brief pause, and so it was time to find another one, and Oh! Oh! Antonio is just what I was looking for.  If you’re wondering why I’m suddenly bringing music hall into this game, after all that Schubert and czardases, I have to point out that rhythmically speaking, behind every balletic variation, there’s a tarty music-hall number dying to show its frilly knickers, and a bit of decorum (which flies out of the window once you put some swing into a waltz) is the only thing that divides these songs from Paquita or Bayadère.  

How I discovered Oh! Oh! Antonio

I wonder what the chances are of anyone knowing this if you played it for class? I didn’t know it until last week, when it was used in Indian Summers, set in 1932:  Cynthia Coffin (Julie Walters) proprietor of the English Club in Simla leads the singing as the local English gather for the club’s re-opening. You never can tell: someone on a forum remembers his great grandfather (Welsh, who spoke no English) singing it to him as a child. Songs have a way of travelling through time. I always thought the Teddy Bears’ Picnic was a song from the 60s (because I heard it as a child) yet the tune was written in 1907, and the words in 1932.

In fact,  this song is also about the way songs get transmitted. You may have noticed that it begins almost note-for-note like Strauss’s Kunstlerleben (Artist’s Life) waltz Op. 316, written in 1867, and regarded as the “twin” of the Blue Danube. Now listen to the words of the second verse:

Her old hurdy-gurdy all day she’d parade
And this she would sing to each tune that it played.

So what you’re hearing are new words to an old tune. But there’s a third temporal layer to this: the third verse (not on the recording, but available here) has the line (just before the chorus)

She faded away, but they say in the streets
The ghost of that girl in Italian repeats…

So this is a song about people talking about the ghost of a girl singing a song that she made up to an old tune playing on her hurdy-gurdy. A Pathé newsreel film clip from 1923 of the “first wireless barrel organ” playing this song adds yet another layer to the story: here is a kind of hurdy gurdy playing, ghost-like through the ether, a song about people talking about the ghost of a girl singing a song that she made up to an old tune playing on her hurdy-gurdy. Incidentally, it’s wonderful to see people waltzing in the street as they hear the music. I nearly wrote “spontaneously waltzing” until I wondered whether perhaps Pathé had placed those people very carefully there to make the clip more interesting. Sadly, there’s no audio on the film.

Reanimating ghosts: songs and musicians

When Florrie Forde sings Oh! Oh! Antonio, she brings that ghost of a song back into the physical present (in 1908, that is).   That’s one of the things we do as musicians  –  let songs breathe a bit longer, or, if you like, plant them in ground where they’ll suddenly flourish again just when they were in danger of expiring. There is no natural process by which “great” songs stay hits purely on their own, it only happens by transmission, and the processes can be unpredictable and strange (and expensive, in the case of Pharrell Williams and Robin Thicke). They are “broadcast” both in the TV/Radio/Internet sense, but also in physical form as sheet music and records. They travel with people as carriers of songs, geographically and temporally.

Last week, I saw a 101-year-old woman teaching Are you lonesome tonight? to a young Filipino nurse in a care home in Streatham. I thought, that’s odd, she would have been 46 when Elvis Presley recorded it – surely this isn’t her generation of songs? But when I looked the song up just now, I discovered that in fact, it was first released in 1927, when she would have been 13 – which makes a lot more sense.

So let’s keep Oh! Oh! Antonio  going a bit longer—why not?  Sing along if you know the words (which you do, because I’ve put them in the score). It would of course be wonderful in any class where there’s an Antonio teaching or dancing, or maybe just for remembering your own Antonio-related history. There’s not an app for that, but there’s a song for it.

Update on 26th April 2015

As proof that the Internet’s power is limited by your diligence and ingenuity as a searcher, I’ve now come across a site that has the original sheet music for Oh Oh Antonio! Fuller and probably easier to read and play than mine, all credit to the person who put this site together.

Update on 31st Jan 2016

…and if you’d like to hear it or play it on the Ukulele, visit Colin Tribe via Youtube