This chart plots 419 performances of the first movement
of the Eroica. 159 performances take the repeat, 260 performances don't. This is over 1/2 of
all performances known to me.
From left to right the recording dates progress from 1924 (Weissmann, Fried) to
2000 (Kakhidze, Abbado).
From bottom to top the average quarter-notes-per-minute goes from 110.74 (Klemperer 70) to 174.58 (Scherchen 58).
If you find this interesting, you might also look here.
red = repeat blue = no repeat
Speculation:
- The consensus "fast" Eroica has remained more or less
constant over the years
(Weissmann, von Schillings, Weingartner, Toscanini, Kleiber, Bernstein, Dohnanyi, Tilson Thomas, Masur).
- The consensus "slow" Eroica expanded steadily from
Pfitzner to Furtwängler to Klemperer, at which point it went out of fashion,
with only a non-consensus "ultra-slow" group continuing downwards
(Giulini,
Celibidache, Kempe). It's been 10 years since an ultra-slow version appeared
-- is it dead or only sleeping?
- After Klemperer, "slow" went back to ~140bpm
(Karajan, Böhm, Jochum, Monteux); but then proceeded to slide as before. At this point
Barenboim is where Furtwängler was in 1950, the lower limit of the consensus.
- It appears that the "ultra-fast" Eroica, which caught on following Norrington 87, has also been slowing down! The
response of the ultra-fast group
(Gardiner, Savall, Zinman, Harnoncourt, Mackerras, Levine) to Norrington seems to have been, "I like it, but maybe not quite
that fast." (Note: Gielen seems consistently fast, like Norrington.)
Miscellaneous observations:
For what it's worth, the
famous "dotted-half=60" mark has been reached -- for at least a few bars in a row -- by Gielen, Harnoncourt, Coates, Norrington,
and Scherchen 58.
The ultra-fast difference: all other performances slow down for certain sections (like the second subject -- yes, even Toscanini and Karajan
do this -- and the first 12 bars of the developement.) The ultra-fast don't. The really speedy ones (Scherchen 58 and Coates) don't even
slow down for the big syncopated 'hammer-blows' preceding the dissonant climax in the development -- even Norrington and Gielen hold back there.
The slowest passages are in performances by Mengelberg and Koussevitzky -- in the midst of basically speedy performances,
they occasionally come to a near dead-stop. Mengelberg gets all the way down to quarter=69 -- dotted-half=23!
Comments please! eroica_project@hotmail.com