Monday, November 18, 2013

800 Words: The Goldberg Variations Ideal

“...two skeletons copulating on a corrugated tin roof.” - Sir Thomas Beecham

It can be done on the harpsichord, or the clavichord, or the clavecin, or the clavicembalo (I still can’t figure out if those four are just the same instrument…), just please, for God’s sake, don’t play it on the fucking piano. The piano is just too boring for Bach. I know, I know, everybody hates the harpsichord, and justifiably so, because the harpsichord sucks. But the instrument to play Bach can’t simply be a harpsichord. It has to be one of those mid-20th century harpsichords big as a battleship with ‘16 stops, pedals, terraced dynamics, and expanded keyboard ranges. The Goldberg Variations require the ecstasy of a church organ and the sensuality of a guitar. The work is actually far more difficult to play on the piano than on the harpsichord, and Bach did not write the Goldberg Variations for a modern Steinway he could never have imagined, nor did he write it for a harpsichord, an instrument whose limitations he was always complaining about, Bach wrote his greatest keyboard work for an ideal keyboard instrument that still doesn’t exist; an instrument which makes the mountains dance with the valleys and rides the music of the spheres to the source of our giant cosmic vibration (wow, have I been taking acid?...). He wrote the Goldberg Variations to be performed on something in between a DIY accordian and a franken-lute.

As I’ve written at length about before, I am an avowed Bach-skeptic. For all his obvious greatness, there are enormous swaths of his music which I find a chore to sit through. I often can’t help holding with Thomas Beecham’s view of Bach: “Too much counterpoint; what is worse, Protestant counterpoint.” Along with all the great beauties in Bach’s music comes the endless note-spinning and prattling on about a God in whom I don’t believe. Considering the seeming universality of his appeal, I can understand why Bach felt justified in writing the way he does. But Bach takes it as a given that you’re going to buy what he’s selling, and he’s usually uninterested in engaging with the concerns of difficult customers. Doubt and human longings can be found aplenty in Handel and Monteverdi, but in Bach you barely find it. Even so, The Goldberg Variations is not just another Bach piece. For me, it’s the ideal house for Bach’s furniture. It has all his weaknesses; there’s plenty of note-spinning in a long-form piece with nary a dissonance to find, but the formal structure is so intensely ironclad that the piece (in a good performance) grows as one listens. Every variation builds upon the last to create an edifice of universal appeal.

But even in this most secular of Bach works, nothing is more important to its vitality than its Lutheran grandeur. Bach may have written this work for the drawing room, but there is little in Bach’s biography to suggest that he was ever comfortable mingling with aristocracy (Handel, on the other hand…). Bach is often thought of as the world’s most practical musician, but he was constantly in trouble with authorities because he could not sufficiently tailor his music to the desires of his intended audiences. Most of The Goldberg Variations sound little different from music he wrote for Church services, and unless the keyboardist can get his instrument to have an organ-like quality to complement the rhythmic vitality which comes so naturally to every keyboard instrument except the organ, the performance will not capture Bach’s particular musical ecstasy. The ecstasy of Bach involves neither the aristocratic sensuality of Monteverdi nor the martial grandeur of Handel, it is an entirely bourgeois ecstasy from the land of Luther. It resounds with the sense of pride which Bach’s contemporaries got from having tilled their fields for 17 hours every day before they spent six hours posing for a Rembrandt painting so they could sleep at night for a half-hour and spend three minutes around midnight impregnating their wives. What Bach expresses is the peculiar Protestant ecstasy of the early reformed Church, in which the possibilities of history were yet again fresh and unspoiled. This is the spiritual need which all Bach performances must convey. There can never be more than a hint of the decadent in Bach, because were there anything decadent about Bach’s music, the early protestants would have repressed it.

There is something about the piano which is simply too louche, too polished, too refulgent, too ‘pretty’ for Bach. The most famous piano recordings of the Goldbergs, (obviously except Glenn Gould’s), are much too conventionally pretty. Murray Perahia, probably more praised in this piece than any pianist since Glenn Gould, has the dubious distinction of making the Goldberg Variations sound like Mozart - everything is limpid and lithe, and much too elegant for a work written by Bach’s organ-sausage fingers. Andras Schiff, their nearest “rival,” is a little better, but still wants everything to be beautiful. There is little more exuberance in his world than in Perahia’s, and the air the performance breaths is vacuum-packed. Wilhelm Kempff plays, as he always does, beautifully, with masterful pacing and great feeling, but where’s the fun? The humor? The high spirits? Simone Dinnerstein plays with assembly line efficiency in the fast sections, just getting through them as quickly as possible so she can bask (beautifully) in the slower, quieter stuff.

And then there are the piano recordings that are just plain bizarre. Rosalyn Tureck recorded it six times, and from what I can bear to listen to, they all sound like Bach preserved in embalming fluid - ultra-reverential, overcontrolled, and slow. By far, the most interesting conception among pianists I’ve heard is Daniel Barenboim’s. As often happens, Barenboim has a musical conception which is leagues ahead of his rivals, but the conception is almost completely wrong for Bach. Rather than dance ecstatically, he tries to coordinate the music to change color with every harmonic shift - but how many colors can you get out of a piece in which the same harmony repeats itself 32 times every 32 bars? Unlike Perahia and Schiff, there’s fun to be had with Danny, but the conception is always more important than the music. Never does Barenboim, however perceptive he is, give himself completely to Bach’s torrential flow of ideas. Jeremy Denk recently wrote a piece for The Guardian in which he claimed that until recently, he hated playing The Goldberg Variations, and his recording makes it sound like he still does.

In many ways, the best piano recordings come from the beginning of the recorded era. No doubt, the pianism of the young Rudolf Serkin and Claudio Arrau was considered the height of austerity in the pre-WWII era because they performed the entire Goldberg Variations in Bach’s own arrangement, but in the hands of these blooming soon-to-be-giants, The Goldberg Variations sounds like a Beethoven sonata or a Liszt showpiece. It’s so unbelievably wrong, and yet so fantastic.

Nearly all the greatest recordings: Wanda Landowska, Helmut Walcha, Igor Kipnis, Karl Richter, George Malcolm, Anthony Newman, were done by harpsichordists who predate the period practice movement’s most rigid dogmas (or, in Newman’s case, was severely  controversial within it). By the time history arrived at historically informed performance ‘Giants’ like Gustav Leonhardt and Trevor Pinnock , the doors of perception were closed, and the Goldberg Variations became just another dusty ‘masterpiece’ which you appreciated without feeling any true passion for it (Ton Koopman is better, more on him in a moment..).

It’s an odd thing about Bach that the two most celebrated players of his keyboard music are a pianist who made the piano sound like a harpsichord, and a harpsichordist who made the harpsichord sound like a piano. Glenn Gould will get his own paragraphs, but at this point in history, it’s almost impossible to talk about Wanda Landowska without talking about Gould too. Gould was a creature of the 20th century, and delighted in Bach the contrapuntalist, the rhythmatizer, the virtuoso. Under Gould’s hands, the sheer overflow of Bach’s invention is enough to make you giddy. If Murray Perahia’s achievement is to make Bach sound like Mozart, then Gould’s great achievement is to make you think that Mozart himself is playing Bach.

But even if she played the harpsichord, Landowska was a creature of the 19th century. She clearly loved Bach the Christian, the celestial harmonist, the life-affirmer. Listening to Landowska at her miraculous best is to listen to the very process of catharsis. Her harpsichord had as many stops on it as some organs, and she was unafraid to use them to create the grandest possible sound. The celestial ‘ring’ of Bach’s musical overtones was never heard more pellucidly than in her performances. Listening to Gould is like listening to delight, but listening to Landowska is like listening to joy. I’m elated and exhausted by the end of either Gould recording, but at the end of either of Landowska’s recordings, I’m consoled and hopeful. And yet even Landowska, this amazing force of nature, didn’t record the greatest harpsichord recording of the Goldberg Variations (more on that in a bit).

Some more iconoclastic players, like Ton Koopman and Anthony Newman, err on the side of too much flash. Both Koopman and Newman (Dutch and American) are somewhat controversial within the ‘early music’ community, both of them are staggeringly virtuosic keyboardists whose performances are so laden with ‘ornamentations’ (improvised extra notes) that they drive other scholars crazy. Next to the forbidding austerity of their authentic performance practice godfather, Gustav Leonhardt, the effervescence of Ton Koopman must have been the most direly needed breath of air in Holland since they switched off of windmill power.  But Koopman rather overdoes it, and his Goldberg Variations feels like a kid on a sugar rush. The spiritual import is gone, and there is nothing in its place. Newman is a somewhat better, but his performance can be downright gawdy. It is an amazingly fun, almost ‘rock’n roll’ performance (pardon the stereotyping…). The virtuosity is so unbelievable that even Gould has to take second place, but intellectually his footing is nowhere near as solid as Gould’s, and therefore it seems a bit like empty (though thoroughly delightful) gymnastics. Both Newman’s and Koopman’s Bach has plenty of personality, but why would you ever want to imprint your personality on a work which is already so perfect?

Other players, more stolid German players like Helmut Walcha and Karl Richter, get a little closer to Bach. In the mid-20th century, there were not two men alive more associated with Bach than these two. Neither was best known as a harpsichordist, but both were known as the ultimate authorities on their repertoire.

(Helmut Walcha playing my other favorite Bach piece…)

Helmut Walcha and Bach organ music are virtually synonymous. No keyboardist has ever achieved easier access to the calm center of Bach’s storm. Every phrase in Walcha’s playing feels so unbelievably right and natural, there are very few ornaments, and every variation has a moderate pulse which stays as close to absolute as humanly conceivable. This is the kind of playing which allows Bach to speak for himself, and no one ‘speaks’ in the language of notes as fluently as Bach, but there’s still something missing. Surely Bach - the great improviser, the world’s most practical musician - wouldn’t stop at a simply perfect rendition of music he already wrote. If he’d already achieved perfection, he would muss it up so he might plumb new depths.

(Karl Richter leading the St. John Passion old school. Sorry, but the St. John Passion kicks St. Matthew’s ass.)

If Walcha was synonymous with Bach and the organ, then Karl Richter was synonymous with Bach and the large ensemble. To be sure, Karl Richter was a great organist in his own right. But he is, by far, best known for conducting the Munich Bach Orchestra and Chorus - a mostly amateur organization which is the only major ensemble in the history of major record labels which I’m convinced comes close to the spirit of Bach’s music. Like Bach on the piano, professional orchestras and singers are far too seasoned, mannered, and lacquered to view Bach on his own terms. Bach went to such artful lengths to craft his works so that his extremely amateurish performers wouldn’t have to perform with any artfulness at all. Bach’s music is the definition of music for skillful amateurs - pure in intent, serious in their passion, and earnest in their feelings. Richter’s performance of the Goldberg Variations transfers this earnest feeling perfectly, and perhaps too perfectly. Until his death from a heart attack in his mid-fifties, Richter was a heavy smoker in terrible health for much of his adult life, and perhaps his performance of the Goldberg Variations is indicative of that. It abounds with wrong notes, slowed up rhythms, and even a bad memory lapse or two. It doesn’t matter much. His feeling for Bach was amazing. Think of it as a fantastic but rusty piano (harpsichord) teacher playing in your living room.

(George Malcolm, the closest to ideal you’ll find on youtube.)

And that leaves two. Two that trump even Wanda Landowska herself. The English keyboardist and conductor, George Malcolm, is one of the forgotten heroes of Baroque music. He, understandably, hated the modern harpsichord, and pledged himself to the newer model harpsichords with modern accoutrements. No harpsichord, not even Landowska’s, sounds like this. In his 1963 recording, is none of the percussive hardness which usually pervades baroque keyboard instruments. The notes on this instrument sound as though they’re being plucked by human finger, in a sound that is the perfect mix of harpsichord, organ, guitar, lute, and piano. Oh… and the interpretation ain’t bad either.

(Igor Kipnis)

But the greatest, closest to ideal of all, is Igor Kipnis. Kipnis didn’t play harpsichord seriously until he was nearly thirty, and while he’s no technical slouch, he does not quite have the easy virtuosity of a Landowska or Anthony Newman. But in its place, Kipnis has that one extra element which no other keyboard player uses to nearly the same extent - improvisation. Bach, who could write a perfectly formed work as easily as the rest of us could sweep a floor, couldn’t possibly be satisfied with a merely perfect rendering of his pieces. In his compositions, he so often mined the same basic musical material for different results that it’s impossible to believe he didn’t do the same as a performer and improviser. And like voodoo magic, Kipnis always has a new way to shock us into hearing differently a piece the world knows so well. He’s never too excessive in his divergences from the score, and were he, it wouldn’t be nearly as shocking. There is just enough diversion from the score to surprise us expertly at every turn. Save one, Igor Kipnis realizes better than any other keyboardist that the score is only the beginning of understanding Bach. And in so doing, he channels the spirit of the master as no other performer ever has in this most Bachian of works.

But what if there are conceptions which are greater than the ideal? What if a musician comes along who completely changes everything we’ve ever known about a work in a manner that could never be duplicated - and turns a work of art into something still better?

This is the accomplishment of Glenn Gould. No work was ever associated with him more than the Goldberg Variations. Under his fingers, the Goldbergs were no longer a forbiddingly austere behemoth, it was a work of pure delight. When Glenn Gould was 23, he set down a version of the Goldbergs so legendary that it’s moreso than the piece itself. Many other players take nearly 100 minutes, but Gould gets through the whole thing in 37! No repeats, and at top speed, with not a single use of the damper pedal! Under every other pianist, The Goldberg Variations is an epic journey, but when it becomes the Gouldberg Variations (sorry), it becomes an exercise of pure counterpoint and exuberance - as though filtered through the mind of a pure genius who assimilates everything at top speed. Murray Perahia makes the work sound like Mozart, but Glenn Gould brings to mind the image of Mozart himself playing it.

But even Gould’s 1955 performance must take a back seat to his 1982 rerecording. Glenn Gould, a few months before his early death, shows us what real genius is when he manages to play the entirety of the Goldberg Variations in mathematically related tempos - sometimes bizarrely slow, sometimes disturbingly fast - but never had this piece ever sounded this beautiful or exciting or unified or masterly under any other pianists hands. Only Gould could do it; is this Bach? Or is it, dare I ask?... an improvement on Bach? Regardless, the result may be the greatest recording of Bach ever made.

No keyboard player could sound like Gould, even if they tried (and many have). Whatever his secret, it’s his alone. Like only the very greatest re-creative artists can, he took the raw material of a piece that was already a work of genius, and transformed it to something even higher. Everything in music is a matter of taste, but there are some classical performances for which the appeal is so obvious and immediate - Pavarotti doing Bel Canto, Furtwangler in Bruckner,  Horowitz playing Liszt, - that if you don’t hear the greatness, I question whether you have a pulse. These moments are more than simply great, they’re moments in which human limitations are clearly transcended in a manner they’ve never been before, and never will be again. None of the above examples, including Gould, are doing precisely what the composer asked, but if anything, they’ve improved on the original. Horowitz completely re-arranged Liszt’s 2nd Hungarian Rhapsody to get those effects, no acceleration is written in the score during the closing pages of Bruckner’s 5th Symphony, and those nine high-C’s which Pavarotti tossed off were originally written with instructions to yodel, but what these performers have done is more than simply realize what the composer wanted - they improved on the composer’s music.  

The two greatest recordings of the Goldberg Variations were, of course, done by a pianist. Glenn Gould is the alpha and omega, the north star by which this piece will always be navigated. Don’t let anyone tell you that his recordings, either of them, are overrated. They are, if anything, still underrated. They are the most profound musings left to us by a genius who uses his imagination to reshape the curvature of the earth to his own design. For whatever reason, the mid-century was full of those artists - Gould, Horowitz, Richter, Cliburn, Dinu Lipatti, William Kapell, Julius Katchen, Leon Fleisher, John Ogdon, Menuhin, Josef Hassid, Michael Rabin, Genette Neveu, Nikolai Golovanov, Sergiu Celibidache, Ferenc Fricsay, Guido Cantelli, Carlos Kleiber, Bernstein, Gershwin, Kurt Weill, Alban Berg, Jacques Brel, Edith Piaff, Jerome Robbins, Bird, Coltrane, Brando, James Dean, Nicholas Ray, Orson Welles, Arthur Penn, Erich von Stroheim, Jean Vigo, Grace Kelly, Carole Lombard, Lauren Becall, Richard Burton, Peter O’Toole, Nicol Williamson, Tennessee Williams, Arthur Miller, Isaac Babel, Vladimir Mayakovsky, Osip Mandelstam, Anna Akhmatova, Mikhail Bulgakov, Boris Pasternak, Paul Celan, Albert Camus, George Orwell, Anne Frank, Sylvia Plath, Richard Hofstadter, Loraine Hansberry, F. Scott Fitzgerald, Margaret Mitchell, Jack Kerouac, Flannery O'Connor, Nathaniel West, J.D. Salinger, Truman Capote, Harper Lee, Norman Mailer, John Kennedy Toole, Joseph Heller, Ralph Ellison, Otis Redding, Hank Williams, Woody Guthrie, Jimi Hendrix, Jim Morrison, Janis Joplin, Sam Cooke, Patsy Cline, Brian Wilson, Bob Marley, John Lennon, Elvis Presley- performers so absolutely gifted that our confused, compromised world had no way of accommodating the true breadth of their talents, and therefore we only have mere glimpses of their genius stretched its fullest plumage.


  1. I agree and disagree in almost equal measure: quite an achievement, though I'm not sure whose...!

    For instance, though I'm really not wild about Gould - I hear the attraction though, and may well come to feel it strongly once again - I agree re 1982 vs 1955. We both recognise Barenboim as being special here, but I don't share your reservations. The St Matthew Passion - well... But then we feel very similarly about Richter. Malcolm: yes, a true musician, appallingly undervalued. (His recordings of Bach concertos with the young Schiff as soloist seem to be better than Schiff's later versions.) I could go on, but wanted to thank you for such a provocative - in the best sense - posting.

  2. Hehe. Barenboim is of course an amazing musician, and his ideas, however I ultimately feel about them, have to be respected. The St. Matthew Passion is odd, I've tried for years, but I don't think I've ever heard an approach that really works. I wonder if a 'sing-along' St. Matthew in which the audience sings along in the chorales would help get the proper atmosphere. Insofar as I can have a favorite recording, it's Richter's (audio recording, not video... that's just too slow). George Malcolm will come back once the pendulum reswings, I'm sure of that. I'm sure you'd love the Walcha recording, but I wonder how you'd feel about Kipnis. Growing up, he was actually a very good friend of Carlos Kleiber's. His recording might strike you as a little too fast and flashy, but I think it's ultimately closest to what I imagine to be Bach's vision, getting both the spirituality and the lightness and holding them in tandem.