Monday, November 1, 2010


Something's been bothering me for a while, and that something is meter. Meter is a fundamental attribute of most music, something so basic we're usually completely unaware of our effortless perception of it. But I get it wrong all the time. Is it just me, or are you all in the same boat? Quick survey: How do you hear the following excerpts? At 0:29 here, is the strong beat on a) the triplet, or b) after? What about 7:20 here (strong beat with a) the winds or b) before) or 2:13 here [beats on a) the first of the three note motif or b) the second]? (This kinda thing happens all over the place in Brahms...please also note the ridiculous-looking, yet absurdly effective conducting in the first link). But it also happens in Bach (are the 8th notes here grouped in a) 3's or b) 2's?) What about this one? Go to the end of Variation 25 (maybe 5:10 on)....where are the downbeats? Or where's the beat in this one [with a) the bass notes or b) just after? Don't peek at the score, not that it will really matter!]. Just a few examples of the thousands available!

So did you botch the listening quiz like I did? (answer b is correct for all the questions). How much did it matter if you've played/seen the scores to the pieces in question? For me, the Brahms excerpts go in and out, often depending on how actively and engagedly I'm listening (see below). But I can't hear Hilary Hahn's Bach "correctly" no matter how hard I try (though I can play the correct auditory image afterward in my head). Ditto with the Schumann.

So maybe the composer's really botched it? Or the performers? Or is it us listeners?

Or is it just me.

For today, two basic questions to consider about meter: how does a listener "decide" what the meter of a piece is, and, delving deeper into the realm of the unexplored and unexplained, what's the role of the performer or the composer in deciding what the "right" meter is, or making sure the audience "gets it"?

Short answer to question #1, which has received a fair amount of research. There are some basic things that signal to us as listeners where the beats are in music, most of which you could probably guess on your own: longer notes, louder notes, and lower notes are all more likely to be found on strong beats, and are taken as cues to those beats by listeners. Articulation and phrasing matter too; we hear phrases starting on stronger beats more easily.

But that can all get pretty boring. Music is at its most rhythmically vibrant (maybe?) when various aspects send us in different metrical directions and leave our final perception on the fence.

That being said, what a composer writes on the page isn't necessarily what we hear, nor is it sometimes even possible to hear it the way it was written. Quick thought experiment: if we scanned a Beethoven symphony into Sibelius and then added a quarter rest at the beginning, therefore shifting every single beat throughout the whole thing, and then played the thing, would it sound any different? Of course not! The score just wouldn't match our perception anymore.

What are we to do when such mismatches arise in real life? First of all, there are different degrees of mismatch between score and perception. Sometimes a remedy may be desirable, and sometimes it may not. But these are the types of questions I rarely hear anyone raise, even when playing Brahms, where these types of mismatches are so common. Is that section of the 3rd symphony, 1st movement "better" if the listener understands that the beginning of the phrases are actually on the 3rd and 6th beats of each bar, not the 1st and 4th? Not sure. How hard should the conductor try to make the audience understand this fact, or should he or she play it as "counter-metrically" as possible (which is more or less how it's written)?

I'm not sure. And incidentally, just "knowing" where the beat is usually isn't enough to change our perception. Even seeing isn't believing (as in a conductor giving a downbeat). But this you can probably figure out introspectively. In order to change one's metric perception, atually moving your head (link forthcoming...too lazy at moment) turns out to be really important, which, incidentally, makes it really easy for the performers to get it right in their minds, but not the audience.

I'm also fascinated as to the extent of these perceptual mismatches, especially in Brahms. 'Cause it really does change how you hear the music to put the beats in different places. It's pretty fundamental. But he makes it so damn confusing.


  1. I remember the first time I PLAYED that thing in Brahms 4 for the first time rather than just listening to it... it seemed totally backwards. Now it sounds much more right, probably since I'm used to playing it that way. There's also a really confusing similar thing in the first movement of Brahms 1. I'm obviously too lazy to look it up right now. This kind of thing happens all over the place in Sibelius (the composer, not the program) too. I've seen conductors just cave in and conduct in the right "sounding" way rather than how it's scored... or, less extreme but still (in my opinion) kind of confusing, advise people to basically ignore the downbeats and count the "right sounding" way to themselves.

  2. Cool trick: if you append #t=7m20s to the end of a youtube url, it'll jump to 7 minutes 20 seconds into the video.

    I don't have much intelligent to add to this discussion, but I'm a huge fan of audtical illusions. Woohoo brahms!

  3. Eli, thank you muchly for the tip, which i have already used. but don't be so conceitedly modest. you know these pieces. how do you hear them? do you find them more rhythmically interesting in their written form or their more naturally occurring aural form? cmaaaan. how willing is your ear to add an extra "beat" to acommodate the "natural" sounding downbeat? these are questions i'm generally curious about. i don't know what the general variance is in how people how people hear this stuff

  4. I'm surprised you think there's a "correct answer" to the questions . . . seems to me the music's ambiguous as to where the beat lies, and that the performer can choose different articulations all of which are 'correct' - no?