The limits of period performance
I don't go on about it, but I really do love Beethoven's fifth piano concerto, which has earned the sobriquet, "Emperor." Of all my recordings, I tend to return to the 1957 collaboration between Otto Klemperer and Claudio Arrau. There is a power and majesty in that performance that, while others come close, none equal. The sound, despite being from 1957, is none of the best, and it is a demerit to Walter Legge that he didn't record the collaborations - which were, more or less, his brainchild. I also like the Fritz Reiner/Van Cliburn record, which was recently re-released on RCA (i.e., Sony) Living Stereo SACD. There are differences in the recordings, but that's another post.
It is one particular recording, the Robert Levin/John Eliot Gardiner period-instrument performance on Archiv, which I want to discuss. Levin plays the fortepiano on this record. Now, generally, I find Gardiner one of the most tolerable period-instrument conductors. His non-period recordings are pretty good, too. His Beethoven symphony cycle (again, on Archiv) is actually very good (if on the fast side). My problem with this recording isn't the musicianship, the sound quality, or even the interpretative stance. It's the damn' fortepiano. Listen to Furtwängler/Fischer, Klemperer/Arrau, and Knappertsbusch/Curzon. Then listen to the tinkling, jangling fortepiano. It doesn't gel. Beethoven's tutti are too powerful in the presence of such an instrument. They'll give a concert grand a run for its money today, even compared to something like the Hammerklavier sonata or the wilder runs in the Waldstein. Gardiner manages to be faithful to a contemporary (i.e., of the composer) concept of the orchestral parts, but things don't work out as well with the soloist.
Now, the argument could be made that the difference between a fortepiano and a full-bore, hell-bent-for-leather Bösendorfer or Steinway shows the cleverness of Beethoven's balance between soloist and tutti. That could be true, I wouldn't dispute it. I would, though, call it an enormous cop-out. Part of the difficulty in bringing concertos off well is constructing and maintaining that balance. Pierre Boulez, in his 1971 account of the 5th piano concerto with Sir Clifford Curzon, which is a very strong second to the Klemperer/Arrau account in my book, managed to create and maintain that balance. He did it by, shock and horror, following Beethoven's score and working to create a seamless whole. In other words, he let Beethoven do the work.
It might seem like I am ragging on one particular recording, and maybe I am. I really like the 5th concerto, and I really think that modern instruments - informed by period practice - are the best vectors for the music. My objections, though, go beyond being annoyed at one performance. Listen to Peter Serkin's rendition of the Hammerklavier sonata. It doesn't, to my ears, have the same sheer power and majesty of a performance by a competent musician (and the Hammerklavier redefines competence, especially in the massive fugue). Indeed, it hardly sounds like the Hammerklavier. Before HIP aficionados get up in arms, let me say that I am very fond of some HIP recordings, like Jordi Savall's hard-to-find Beethoven 3rd or Gardiner's Matthäus-Passion (when I don't listen to Klemperer's), but I think that there are limits to the genre.
I'll be blunt. Do I really care how an 1805 audience heard the Eroica? No. Music, either compositionally or performance-wise, did not stop in 1805. Other, equally valid, performance traditions sprang up between a given work's premiere and 2008. That's the thing, when you choose McCreesh's Matthäus-Passion over Klemperer's or Gardiner's 5th piano concerto over Knappertsbusch's with Curzon, you are essentially discarding another performance tradition in favor of a newer-older one. There are limits, indeed, to period-performance, and they start when a work has a performance tradition of its own, and they go so far as to include a performance of a work that sounds "off" or less-true to its intentions on period instruments.
A rant, true. A rant, though, that has been a long time coming.