It has to do with this article by Bernard Holland of the New York Times, in which he ruminates on the nature of ownership and interpretation of a work of art, and whether a composer's performance of a work should necessarily be regarded as Gospel. The artifact in question is CD of Debussy playing his own works, as preserved through the remarkable medium of the Welte-Mignon player piano, which recorded not only the basic performance elements of pitch in rhythm in real time, but dynamics, articulation and pedaling as well. This recording resurfaces from time to time in various forms, and I was thrilled to discover that it's been reissued by the Pierian Recording Society. It was out of print until now, I'd lost my tape of it, and the single copy owned by the Sibelius Academy where I teach was recently stolen. I was able to order a new copy, and plan on keeping it in a safe location, because this recording is something akin to a holy relic to me, a sort of Grail. Here's why:
Years ago, back when I was an ignorant youth just starting to study music, indeed, just learning to read music, I had the good fortune to get into Vanier College in Montréal, a small school in Québec's CEGEP system (a two-year pre-University program something like a junior college in the U.S.). In my second year, after passing basic keyboard techniques, I had to take private lessons in piano as a second instrument, for which I was assigned to Earl Wilson, as iconoclastic a teacher as one could hope to find. We immediately hit it off, being of similar temperament – Earl remains a close friend to this day. However, I made scant progress at first. After listening to me plonk my way halfheartedly through my poorly prepared Clementi and Kabalevski sonatinas for a few weeks, Earl correctly – and very wisely – deduced that I was more interested in learning about music than in being a pianist. At my next lesson, he brought out two scores, Schubert's Impromptus and Debussy's second volume of Preludes, played through a few and told me to pick one of each.
That's when my illicit affair with the piano began, stealing practice time from the horn, keeping me at school until the late hours trying to get my uncooperative fingers around music that was much too difficult for me, but oh-so-much-more rewarding to play than anything I'd been exposed to previously. (For the record, I chose the Impromptu in A-flat and Canopes). As I slowly got the hang of both pieces, the one that really fired my imagination was the Debussy, with its weird ninths and tenths, parallel chords and disconnected registers. That little piece hit me in a way I hadn't expected, and I started really paying attention to 20th-century music for the first time. Then I heard La Cathédrale Engloutie for the first time in music history class, and that was it for me. I remember being completely transfixed by its luminous soundworld, all those ringing, clanging chords and resonant textures, and became obsessed with wanting to play it, but didn't dare ask for more until I'd mastered the first two pieces I'd been given. At Christmas break, once I'd successfully performed Canopes, I hesitantly asked if my next project could be La Cathédrale, and to my surprise, Earl said, "Sure, go ahead."
So I got a copy and started working over the vacation period, staying up until 4 or 5 in the morning to practice. By the time school started again, I'd learned most of the notes, but something didn't fit. I couldn't figure out why the form of the piece didn't make sense when I played it. It seemed that just as the piece was beginning, after the first six bars it would come to an abrupt stop, just floating there with no momentum. I respected Debussy's rhythms fanatically, but I still couldn't make the transition from the opening gesture to the second idea work. The climax felt way too slow as well. I talked it over with Earl, and he suggested doing a lot of comparative listening with different performances. So I hauled myself downtown to McGill University's music library and started pulling out records. It seemed like everyone had a different approach to the rhythmic problem, but the end result was the same: sections seemed too slow, the stasis that took over the piece just wasn't right – even though, as people who know me can attest, I love me some stasis in music. Furthermore, why was La Cathédrale so much longer than the other preludes? It was clearly the climax of the set, but its length was so out of balance with the others that it almost pushed them out of the way.
After trawling through the depths of the card catalogue, I came across a curious entry: Debussy plays Debussy. What!? Why hadn't I seen this earlier? The record collection in question was very rare and kept under lock and key, it turned out, and without a McGill I.D. card, I wasn't supposed to have access. But after much wheedling and begging with the desk clerk, the ancient red-leather bound folder was mine for an hour. I carefully placed the record on the turntable, brushed it off, and started it up. All of a sudden, everything made sense. The mystery was cleared up. It was like a divine revelation, like being given incontrovertible proof that God exists. And this is where we come back to Holland's article. As he writes:
In slow pieces like “Danseuses de Delphe,” Debussy the interpreter confirms the elegant classicism of his score-writing: simple movement without gimmicky flourishes, rhythm observed with a dignified precision.
When the tempos go up, so too do the pianist’s impetuosity levels. At one point in “La Cathédrale Engloutie,” Debussy nearly doubles the written tempo in relation to the note values around it.
It's the second sentence that bugs me. Holland is right at the beginning of the quote: rhythm and tempo in Debussy are extremely important, and I often feel that performers of Debussy, expecially in the Preludes, go for rich sound and flashy technique at the expense of the rhythmic structure. Composers can be notoriously unreliable interpreters of their own music (or at least their own scores), especially if their instrumental technique isn't up to the job, and sometimes notation is just an approximation of the image a composer has in mind. One has reason to ask, as Holland does, whether the composer's own interpretation is what they really meant or not. But there is evidence to suggest that Debussy was no mean pianist, and a reliable interpreter of his own solo works and songs.
So saying that Debussy gets carried away with the tempo in La Cathédrale, a piece whose challenges are more of an interpretational than technical nature, is oversimplifying one of the most ambiguous issues in Debussy's output: the metrical indication at the beginning of the piece. Debussy gives the meter as 6/4=3/2. What does that mean? On the surface, it seems to be a phrasing indication, drawing attention to the way the music veers back and forth between quarter-notes and half-notes, and the necessity of adjusting one's phrasing in consequence: phrases of six quarters versus phrases of three halves.
But is that really it? (It's not a rhetorical question; I ask this in my analysis class every year, and it usually sparks a long discussion, especially if there are more than a couple of pianists in the room.) I don't think so, and closer attention paid to Debussy's performance shows why. It's not that Debussy gets carried away by his fancy, or some other cliché of the performance experience, deciding on a whim to really accelerate the piece. After listening through it a couple of times all those years ago, listening to the composer play consistently faster in sections that were usually much slower, it hit me what that meter meant: that phrases in half-notes were meant to be played at twice the speed of the quarter-note tempo. Hearing Debussy's interpretation, it all fell into place. The massive, arch-like structure of the piece was revealed, tracing a clear rhythmic line from beginning to end. No more dead sections where the momentum fell off. The bell-like melody starting in the seventh bar came alive, where before it had just drifted around. The climax was glorious instead of ponderous. The end rippled instead of sloshing.
I played that record over and over again, as many times as I could in my allotted hour, and went back to the library several more times while I was learning the piece. Earl agreed that it was the best possible solution to the problem, and made the piece stronger. Subsequent piano teachers were also eventually swayed to this new (or old) point of view. I've played it this way ever since, and the piece keeps getting deeper and more mysterious. Something about La Cathédrale, and hearing Debussy's take on it, invaded my consciousness, obliterating all the other music I'd heard until that point, and setting the bar for everything else that was to come. I wanted nothing more than to write music that would rise to its level, or at least come close. It wouldn't be until several years later that I'd be ready to make the commitment to being a composer full-time, but Debussy showed me the way. If I hadn't heard Debussy's "ghost", as Holland puts it, in that recording, would I be here? Difficult to say, but I like to think that's where it started. So regardless of whether Holland thinks that Debussy's performance is authoritative or merely whimsical, for me it stands as the authentic, most insightful performance of La Cathédrale Engloutie, the one from which all others should flow, and the best way into a real understanding of Debussy's music. I can't wait to get my own copy, at long last. If I'm very lucky, I'll be able to recapture some of the spark of that original revelation.