Piano World Home Page Forums Our Most Popular Forums Pianist Corner Debussy - Poissons d'or - Rhythm?

I'm working on Debussy's second book of Images, and there's a troubling pair of bars in Poissons d'or.

You can see the two bars at the bottom of the image below. Note that the piece is in 3/4 time.



Any thoughts on how to fix the rhythm? The bottom two staves imply a bar with 4 beats, though the top staff has only 3. Note that the rhythm in the right hand is a continuation of the motif established previously.

This snippet is from the Durand edition, though my Henle is more or less identical.

Any thoughts?

Count the downbeat 1. The LH G# (the second note) comes in on 'and' and the second RH chord on 'a' (along with the LH rest).

The LH flourishes basically needs to be about as fast as you can play it.
_______


The low C# should be written with 2 note heads. A dotted half but also and 8th for the run. This is the way Henle represents it.

Very interesting!

It's part math problem, part music logic, isn't it....

And you're right -- taken literally, the lower 2 staves look like 4/4 time.

So.....what to do with it.....

My first premise is, the rhythm of the upper staff is exactly what it looks like, and the basic 3/4 is maintained.

Which means that everything else has to fit within that.

So......looking at that first measure....

The lower 2 staves are clearly indicated as being associated with the upper staff in a very specific way. The 2 quarter-rests correspond to the 4 eighth-notes above them. Those quarter-rests are just what they look like: each has the length of a quarter note.

That leaves us having to figure out what to do with that low note and the couple of riffs in the first part of that measure.

I think we have no choice but to conclude that the indicated note values either:

-- weren't meant literally, or
-- Debussy or the editor or typesetter just made a mistake on them.

Be that as it may, whichever (doesn't matter)......

The low note plus the quintuplet occupy the first half of the first beat, i.e. a total duration of an 8th note.
The next figure plus the 16th-rest also occupy the length of an 8th note.

I think that summarizes the heavy lifting on this. The rest is just scribbling. (As Mozart says in Amadeus.)

In order to make this 'work,' we have to either not worry about the indicated note values, or to imagine that they're written double too long -- i.e. the 8th-note stem on the low note should be a 16th note stem, the 32nd notes should be 64th's, etc.

There are other slight complications in the decisions, such as how to regard the quintuplet, and what to do with the second measure, but I think once we have these basics in mind, those are relatively simple and you can figure them out (as could anyone else who's advanced enough to be really delving into this).

So -- whaddaya think?


edit: I see that Dan covered this too, and basically the same, only with a lot less words.

This is a very reasonable possibility. I had always played this under the idea that the 5 and ensuing LH notes were supposed to be twice as fast, but I never considered that the low C# might be a 16th. That would allow for a nice 16th pulse leading up to the 2nd melody chord...nice! Either way, the important thing is to keep that melody in time!

Another way of saying what DanS wrote earlier and what Marc_C has written :

In spite of the "stemming" of the LH notes, it seems pretty logical (obvious?) to me to assume that :

- the low C#
- the next group of 5 notes
- the second group of 5 notes

are each to be played in the time of a 16th note, given that there is a 16th rest following and under the last 16th in the first beat of the right hand. Thus, there is the value of four sixteenth-notes in the LH while the RH plays a dotted eighth plus a sixteenth-note. Both hands thus give four sixteenth-notes making up the first of three beats in 3/4 time.

We are, I believe, essentially all in agreement here.

Regards,

Hey everybody,

Thanks for all of your responses. The idea that everything should be crammed into a single quarter beat seems like the best idea, though it's difficult because the tempo is already quick enough. For what it's worth, Michelangeli uses two hands for the arpeggio and plays them as an almost unmeasured blur, but the right hand rhythm gets stretched just a bit. Aimard and Rubinstein treats the rhythm as written, but continue the arpeggio into the first quarter rest, so everything fits.

I'll see how fast I can play the arpeggios, but if nothing works I'll just play into one of the rests.

While I'm not sure because I haven't played the piece (and I'm too lazy to take this over to the piano and try it) ....I would think that those figures could be played at almost literally infinite speed, without limitation.

I wouldn't even call them arpeggios, because I view them more as rolled chords.
You can roll rolled-chords as fast as you want, without limitation -- right?

Scriabin has lots of stuff like this. I have basically no technique at all .....but I can play those kinds of figures at whatever speed you might want, just by seeing them as rolled chords.

I'm guessing that this is exactly what Michelangeli did -- and it's why he used two hands.

I would think that whatever rhythm fish have is not very precise.

This doesn't count, of course:

This is true, but I like the smooth washing effect of measured arpeggios, which I think is also appropriate for a water-like piece. A rolled-chord solution works, but then the delay from thumb-turn starts to stick out... (also an issue with those big rolled chords in Cloches a travers les feuillies). You're right though, using two hands like Michelangeli would also work.

No, because:


It can only be done the way I said by using two hands. I meant using two hands when I said "rolled chords."