Meter vs subdivision: An easy solution to the ballet and time signature problem?


In the course of writing an article about the problems of ballet and time signature recently, I read a lot of books and articles about 18th century theories of meter. I’ve already blogged about that several times recently, as I found a lot of it helped to answer queries I’ve had for years.

As the dust has settled on the thoughts that I had for the article, I have come to a conclusion, for now at least, that “mixed meter” as the 18th century theorist Koch used that term, offers a really useful way of understanding meter and time signature problems in ballet teaching. I began to formulate that idea in And now for something completely sextuplebut I could say it even more simply now.

The problems I’m talking about are almost exclusively to do with triple meter, and the fact that in the kind of music that you get in ballet classes (i.e. predominantly 8-count or 8-bar phrases) there is always a possibility of counting in 2 or 4 rather than 3 – and if trainee ballet teachers have to learn about time signature, they often have difficulty understanding why something should be classified as a 3 or a 6 when it seems more natural to count it in 4.

If Koch were to come back to earth and teach such courses, he probably  wouldn’t say  “you’re wrong, it’s a 6,” he would have said “this is a tripled 2,” or “mixed meter” – you’re right, it is four, because the tune is moving predominantly in a 1,2, 1,2, fashion, rather than diddeley diddeley diddely diddely style, like a double jig, or Schumann’s Der wilde Reiter/The Wild Horseman.  Koch’s point was that these things that sound like fours, are fours, but you write them in compound time like 6/8 because it’s notationally easier. A true 6/8, or truly triple meter, is one where the movement of the tune coincides with the “6” in the top of the time signature.  The exceptions – and they have to be learned – are the triple jig, the polonaise, and the sarabande, and very slow versions of either minuet or mazurka (and of course, many other styles of music that are not so commonly used in ballet).

To put it another way, what matters about time signature if you’re trying to make a bridge between movement and music is not how many beats there are in a bar, but which beats carry the motion. This rather 18th century way of looking at it helps to distinguish meter from subdivision. If you define time signature only as how many beats there are in a bar, without also looking in individual cases at what  happens  in the melody or most salient part of the musical surface, then meter and subdivision get confused. That’s the kind of class where hapless students try to count “12” in a bar of 12/8 where Koch would have said, “Don’t count 12, because this is mixed meter, it’s really a 4, written in 12 for the sake of notational simplicity.”

It’s rather a shame that I’ve only just realised this at the point in my life where I no longer have to teach it, but that’s why partly why I keep a blog – to atone for my past conceptual sins.

Print Friendly, PDF & Email

2 thoughts on “Meter vs subdivision: An easy solution to the ballet and time signature problem?

  1. Chrissie Brown

    I so agree! I’ve been trying to teach basic music theory to ISTD DDi and DDE students for years, and over and over again I find myself saying something along the lines of “Look, don’t worry about it. Just listen to your body and sing to me what your body is saying. Let me worry about the time signature.” Because, yes, absolutely, very often the question is decided by what is easiest to notate and hence to read. I’m a bit depressed by the current situation where composers are obviously using software which prints the notes but not necessarily in a way which is easy to read.

    1. Jonathan Still Post author

      It’s an uphill struggle isn’t it? And since I read the book by Roger Grant (see my latest post) I’ve realised that the problem is also that almost everyone thinks of time signature as being static, as something you can just bark out as an instruction as if it’s information, rather than something that has implications of movement. Big topic!

      I agree with the notation issue too: in years of putting stuff into Sibelius from old scores, I’ve realised how much people in former times used notation to help people make sense of the music (like beaming notes together as a form of phrasing), but you have to spend a lot of time thinking about it to make those kinds of details, and a lot of it just comes spewing out of the computer unedited these days.


Leave a Reply