In my last post on Beethoven 3, I talked just a little bit about my experience of studying the work with my teacher, Gerhard Samuel (more on him here and here). It remains a particularly memorable experience since it was the first score we went through together, so I much of what I learned about his ways of looking at a score I learned through our work on this piece.
Not all went smoothly in those sessions. Like many musicians and listeners, I struggled sometimes to understand Beethoven’s tempo indications. Gerhard’s outlook on tempi was very simple- one took the composer’s tempo, whether you understood it or not. “You don’t need to understand Beethoven- you just need to respect what he’s written.” I, on the other hand, felt, perhaps with an exaggerated sense of self-importance, that I had to understand why that tempo was right, and that I had to believe in it.
In particular, I couldn’t seem to accept what seemed like a perversely fast metronome marking of 8th=80 in the Marcia funebre. I’d grown up with some monumental performances at tempos almost half that speed by people like Furtwangler and Jochum. Of course, when I started conducting it like that, Gerhard screamed like I’d just closed his fingers in a car door and told me I was far too slow. “But it’s a funeral march! It is supposed to be slow!” I protested. I clumsily tried to demonstrate my idea of a funeral march while seated…
Whereupon, Gerhard stood up and showed me the actual march steps of a traditional Austro-German funeral march, while singing the Beethoven. Around the studio he walked- it’s a very distinct pattern of steps, and would look quite absurd at the slower tempo.
About ten years later, I conducted a performance of it here in the UK that got filmed. One of the violinists in the orchestra took the DVD with him to Germany (unbeknownst to me) to show his friend, an elderly but very distinguished critic and Beethoven scholar. My colleague gleefully reported his friend’s reaction to the slow movement.
“Jah! Dis is a verry gut tempo- ein nice balanz betveen zee strong und veek beats, und die tempo is sehr gut for zee real marcia, ven vee valk like dees..” Whereupon, the great man stood up from his couch and started marching around his study to the strains of our recording, while my friend stared on in amazement. My friend’s recreation of the great man’s march looked almost exactly the same as what Gerhard (who was, of course, raised in the same part of the world) had showed me ten years before.
That the same passage of music should inspire the same “dance steps” from two men in the know, 5000 miles and ten years apart is not really surprising. This kind of march would have been a well known civic phenomenon, which was a part of the culture for many generations. It’s no accident that the opening Trauermarche of Mahler’s 5 Symphony is all built of the same rhythmic materials (albeit notated in a different meter- in Beethoven, the eighth note is the pulse, while in Mahler, it is the half note- the tempo is almost exactly the same). There are two good reasons for this- Beethoven’s Marcia funebre was Mahler’s obvious model, and, Mahler was also looking to the same models in vernacular music as Beethoven.
In both the case of the Beethoven and the Mahler, the music’s roots in vernacular music have led the composer to try to find ways o notating something that is not easily written down- the speed of the triplet pickups. In Mahler’s case, he puts a footnote in the score suggesting that the triplets should be played “flüchtig” or faster than marked. The easiest “cheat” for this is to play the figure in 6/4 instead of 2/2- by starting the triplets on the 3rd subdivision of a beat divided in 3 rather than in the middle of the beat, you get something much closer to the military fanfare Mahler is asking for. However, Mahler also says the triplets should be played “quasi accel.” Taken literally, that would mean that that the triplet starts a bit slow and finishes a bit fast, relative to even the 6/4 subdivision. It’s an effect that is hard to describe, but easy to recognize if done correctly.
Beethoven also tried to get across the un-notatable nature of the triplet rhythm in two different ways. First, it is helpful to remember that “Marcia funebre” is not just a poetic title, but a character and tempo marking. It is likely that any musician of Beethoven’s day would have seen that indication alone and known the vernacular models Beethoven had in mind, much as we know that “Tempo di valse” doesn’t just mean “in the tempo of a waltz” but really means “in the rhythmic language, musical style and tempo of a waltz.”
To hammer home his point, Beethoven first marks the triplets in the double basses as grace notes. This has caused generations of confusion because apparently there are some easily addled conductors our there (is there any other kind?) who think these grace notes should be played on, rather than before the beat. The sheer lunacy of this notion is quickly dispelled by comparing the notation at the beginning to that at bar 6, when the figure returns (lunacy or not, it’s common enough that every principal bass I’ve done the piece with has asked me about it, which means they’ve worked with at least one conductor who wanted them wrong, er, I mean on).
So, why not just mark it the same way at the beginning and at bar 6? Why not just play the opening grace notes as a triplet, the way it is notated- surely that is more accurate?
Well, it’s not more accurate if this is really a “Marcia funebre.” If Mahler’s description of the rhythmic language of a traditional funeral march is to be trusted, then the grace notes are a more accurate representation of what Beethoven has in mind!
So why change to written out triplets? Well, long story short, while it was perfectly legal in those times to write changing pitches as gracenotes, but to do so on a repeated pitch (as all the strings have it at the end of bar 8 ) was verboten in those days. Again, it seems likely that Beethoven’s contemporaries would have known that the triplets should be played slightly faster than written, quasi accel, like a military fanfare or the roll of a military drum. Beethoven scholar Clive Brown has said in several essays that the modern obsession with absolutely mathematically precise realizations of Beethoven’s rhythms would have been totally alien to the performing practice of Beethoven’s day.
The big difference between the scores of the Beethoven and the Mahler is that Beethoven doesn’t tell the performer “how” to perform the music- he tells the interpreter how it should sound to the listener (this is something I’ve riffed on many times here). Mahler is much more explicit about using the score to make clear exactly how the performer should go about getting that concept across. When actually first composing his works, Mahler’s approach to notation tended to be much closer to that of Beethoven- it was only later that he would add all those famous dynamic nuances and footnotes. Compare the first page of Mahler’s autograph of the 5th:
With the final, revised version of the score:
Beethoven assumes a much greater level of intuition and expertise from his performers than Mahler does, but of course Mahler was writing for an international audience, and for the audience of future generations. While Beethoven could assume a high level of familiarity with the musical nuances of an Austrian funeral march, Mahler realized that listeners and performers in distant lands would not know a language that was already dying in the German speaking world.
As a final exercise, I thought it would be fun to mark up the first page of the Beethoven as if it were by Mahler. The point is not to come up with a document that would lead to the performance of Beethoven in a Mahlerian style, but to show what Beethoven’s music might look like in modern notation.: