In the past couple of days, I've been seriously considering what I want the final tempo of the 2nd movement of Rachmaninov's second piano concerto to be. The middle movement of Rach 2 has several quite drastic tempo changes. To me, these are extremely important. When I started learning the fastest, and possibly most technically challenging, section of this movement (right in the middle, leading up to the cadenza - marked 'Piu Mosso'), I had set a target speed on my metronome which, at the time, I thought the fastest REALISTIC speed I could achieve.
As it turned out, by the time I'd worked that section up to my target speed of crotchet = 70 bpm, I felt as if I could easily do it a bit faster. So I set a new target speed: 80 bpm.
I'd already made the decision that I wanted to take this movement very slowly, even in the faster sections. I wanted all the notes to be clearly audible, all the wonderful and bizarre chromatic harmonies to stand out. But once I'd perfected the 80 bpm speed, I suddenly realised that when I stopped playing with the metronome, the technical aspect was so effortless that I instantly sped up.
On all the recordings I've listened to the 'Piu Mosso' section is played incredibly lightly, nimbly, and the final part of it sounds almost more like a trill than proper passagework. When left to my own devices, free of the constraining metronome, I
couldn't help but imitate this speediness now that I had the technique
Of course, different pianists DO take it at different tempos. I even went to the trouble of working out the tempos of my 3 favorite recordings: Rachmaninov took the fast section at a whopping 125 bpm, Yefim Bronfman at 115 bpm, and Van Cliburn (whom I admire specially for his slower interpretation) at 100 bpm.
Yet all of these pianist play the fast section much faster than I was aspiring to.
In my practice session tonight, I tried a number of different metronome speeds, ranging from my original target of 80 bpm to Van Cliburn's nimbler 100 bpm. I couldn't make up my mind which tempo was best. On the one hand, I wanted all the notes to be audible, and not to simply sound like a blur. On the other hand, it felt so much better to play fast, and this was my natural tendency as soon as I turned off the metronome. So I did a little experiment.
I played from a few bars before the start of the fast section at a tempo I found comfortable and thought sounded nice. When I reached the fast section, I let myself settle into the tempo that felt natural, and which I always reverted to when I wasn't consciously aiming for a particular speed.
Then I stopped and turned on the metronome, adjusting the tempo until it matched what I'd just been playing. It was 90 bpm - halfway between my target and Van Cliburn's tempo.
The issue of tempo choice is more complicated than just choosing an appropriate tempo for the fast section, however. There are, as I have already mentioned, several other tempo changes in this movement, and how they all relate to each other is vitally important. If I decided to change my target speed for one section, I have to adjust all the others to fit in with it - it's like a sort of tempo 'ratio'.
For now, I've decided that 90 bpm is my tempo of choice for the 'Piu Mosso' section, but the process has made me realise that I have to be careful about sacrificing my musical intentions for the sake of virtuosity. I'm not really interested in technique for the sake of technique. One needs technique to execute one's interpretation of a piece, but if the interpretation doesn't demand virtuosity, one has to know when to hold back.
P.S. On the spur of the moment I decided to video myself during my practice session this evening, which means I can now actually hear how the different tempos I was trying out affect the sound of the music and my degree of accuracy and so forth. I've never done this before, and it was very interesting. Here's a link to the video