Rhythmic ambiguity in the Bach E major partita prelude

I have been creating a series of beat-driven remixes of canonical classical works. I have mostly done this for my own enjoyment, because I like hearing the pieces with some groove to them. But I also sense that there might be pedagogical applications for this method as well. I finally found a good example: the rhythm in the bariolage passage from measures 17-28 of the prelude to Bach’s violin partita in E major. Listen to it at 0:24 in Viktoria Mullova’s recording, it’s the purple part:

Something strange happens whenever I listen to this passage: after the second measure, I start hearing the rhythm wrong. I bet you do too!

The passage is made of four-note groups. The lowest note in each group jumps out at you as being the most prominent one. They are in a different register from the others, and they define the harmony. You start hearing these standout notes as being accented, even if the performer isn’t accenting them. The convention in classical music is to put accented notes on strong beats. So you probably start hearing the lowest notes in each group as “downbeats,” and your sense of the meter reorients accordingly. But this is wrong! Each low note falls on the last sixteenth note of each grouping, not the first. You aren’t expecting such hip syncopation in 18th century music, so when the passage ends you get all confused about where the beat is.

Here’s the score, with the “accented” notes in red. If you are anything like me, you will quickly fall into a groove of hearing the red notes as downbeats beats, so the last note will feel strangely misplaced.

Apparently, Bach is too clever not just for me, but for everyone. Most performers of the prelude accent the low notes, or at least neglect to accent the actual downbeats. This makes the perceived rhythm of the passage re-orient around the fake downbeats. When you get to measure 29, there’s an awkward reset when the accents go back to where they normally would be. Less-skilled performers stagger straight from one rhythmic feel to the other. Better ones (like Viktoria Mullova) disguise the reset by slowing down in measure 28 and making you briefly lose your sense of the beat. But the reset is still a discontinuity that jars against the flow of Bach’s perpetual motion writing.

It isn’t the performers’ fault that we hear this wrong. Even when listening to the robotic sameness of MIDI or Noteflight playback, I still accent the low notes in my head. There’s just no way to hear the passage as Bach wrote it… or so I thought. Then I tried listening to it over a beat. Suddenly, the whole rhythm reoriented correctly in my head.

It sounds even better with Viktoria Mullova over beats by Doug E Fresh. Listen at 0:28.

Heard this way, the passage makes perfect sense as Bach wrote it. It’s so obvious when there are drums underneath! (It also helps to hear it at a slower tempo with quantized rhythms.) Having trained my ears on the remix, I’m now able to effortlessly play it on guitar with the syncopated rhythm that Bach wrote. I’m wondering what other rhythmic mysteries in the work of Bach or other composers might become clear with the addition of Afrodiasporic beats. Leave your suggestions in the comments.

Update: I just found out that David Bruce made a great video about this piece too:

Hear all of my classical remixes here.