Not many people know who Johann Gottlieb Goldberg was, but plenty of people love Johann Sebastian Bach's Goldberg Variations, and the name is no coincidence. Goldberg (born in Danzig in March 1727) was a talented harpsichordist and composer who studied with Bach's son Wilhelm Friedemann Bach and then with J.S. in 1742-43.
The story of Bach's Goldberg Variations goes that a year or two before Goldberg studied with J.S., Bach was asked to compose a harpsichord piece with which the youngster could entertain his patron, Count Hermann Carl von Keyserlingk, a Russian ambassador to the Saxon court. Goldberg had come to Dresden with the count in 1737, and the insomnia-afflicted von Keyserlingk frequently had Goldberg divert him with harpsichord performances during sleepless nights. This tale dates back to 1802 and the very first biography of J.S. Bach, written by Johann Nikolaus Forkel based on conversations with Bach's son Carl Philipp Emanuel Bach, but there have been doubts cast on this story. When J.S. Bach published the Variations as the fourth volume of his pedagogically oriented Clavier-Übung in 1741 or '42, not only was Goldberg's name not in the title, Aria with 30 Variations, neither was there a dedication either to him or to von Keyserlingk. Since the count supposedly presented the composer with a golden cup filled with 100 gold coins in gratitude for the composition, the sort of generous gesture that, especially back then, tended to get one's name attached to a piece on publication, there seems to be something wrong with the tale.
However, it's easily possible that Goldberg did play the piece for von Keyserlingk, even perhaps with some regularity, and that the count expressed his gratitude. Certainly one of the most common objections -- that a 14-year-old youth wouldn't have been able to play such a complex and challenging masterpiece -- is unfair to Goldberg. By prodigy standards, 14 isn't really all that young! (In 1994, Russian pianist Konstantin Lifschitz made a spectacular, if somewhat eccentric, recording of this work for Denon when he was 17 -- and due to its contrapuntal cross-hands passages, the piece is harder to play on the piano than on the two-manual harpsichord for which it was written.)
There is no question that the 1955 recording by Glenn Gould sparked the popularity of the Goldberg Variations with present-day listeners. It was his debut album, and an extremely unusual choice in those days when pianists made their reputations almost solely with 19th century repertoire and Bach was barely played. It was mostly harpsichordists, to whom mainstream listeners paid little heed anyway -- aside from, arguably, Wanda Landowska, who deserves credit for bringing the piece to public attention with its first recording (1933*) -- who were playing the Goldberg Variations at that time. (*However, Rudolf Serkin made a piano roll in 1928, and that was issued decades later as a recording on the Archiphon label.)
The flair with which Gould played them was an entirely new matter, however, and captured quite a bit of attention; the LP nudged a Louis Armstrong album from the top of the chart. Gould's precise technique allowed him to clearly separate the strands of counterpoint (additionally, he used very little sustain pedal, also unusual for pianists then), so, far from being novel passing fad, his recording still holds up 50 years later, despite the pianist's mildly obtrusive humming. There are actually four Gould recordings of the Goldbergs. The first came in 1954 for the Canadian Broadcasting Company; in poor sound, it's mostly of historical interest to Gould fans. Gould's last recording, for Columbia in the studio, was a year before his death at age 50 in 1982. Always interested in the technical aspects of recording, he wanted to use then-new digital sound, and he had also rethought his interpretation of the Variations in the intervening quarter-century. The fleetly spectacular playing of 1955, with no repeats, gives way to a more contemplative approach in some movements, notably a more emotionally mature and legato 25th variation, and some repeats are taken. The best Gould recording, at least from an interpretive standpoint, came in 1959 at the Salzburg Festival. Notoriously perfectionist, and so uncomfortable with the spectacle of concerts that he retired from public performance in 1964, Gould would likely disagree. Nonetheless, the occasional finger slips that briefly mar a few moments are of little note compared to the greater communicativeness and emotionality on display.
For a long time, Gould was felt to "own" the piece as far as piano versions were concerned; the only debate was whether 1955, 1959, or 1981 was best. But new contenders have emerged in the past decade, and in my opinion he has been dethroned by Murray Perahia (who, by the way, also believes that the 1959 Salzburg concert reading was Gould's best recorded Goldberg Variations). In 2000, a year when Koroliov and Hewitt released admirable piano traversals of this work, Perahia easily trumped them with his stunningly fresh reading. It's not just that his pearly tone is unmatched, but also because he's done his Baroque homework. He not only plays all repeats, he usually varies the ornamentation the second time through, sometimes quite considerably -- more than some harpsichordists, and more tastefully and thrillingly than many. His understanding of the dance rhythms at the core of many movements is acute and multi-faceted, making this technically dazzling performance both faster and livelier than most piano competition since Gould in the 1950s, yet also emotionally stirring. He also carefully structures the overall shape of the piece -- which takes much more than just deciding what to do with repeats. Perahia really takes care to relate the movements to each other dramatically. This is a performance for the ages.
Hewitt and Koroliov are worth considering as alternatives. Angela Hewitt is a Bach specialist who has recorded a great deal of his keyboard music for Hyperion (basically, all the harpsichord works); her Goldbergs may be the most charming and carefree piano recording. She uses the sustain pedal sparingly, with great clarity of line but no sacrifice of legato tone when appropriate. She never lets the rhythm become stiff or overly regular, but generally keeps agogic hesitations moderate and unobtrusive. Each variation has its own characterization. She generally opts for gentle lyricism where justified (and brings out the pathos of Variation 25 without exaggerating it), aided greatly by her supernal command of timbre. However, she revs up the intensity when called for: Variation 16, for example, is boldly French without being hectoring or distorted. In terms of digital technique, Hewitt is nearly impeccable. None of the cross-hand passages cause her the least trouble; only a slight unevenness on left-hand trills could be criticized. Anyone who has played this music will be dazzled by the seeming ease with which she tosses off its many challenges; rarely has this work sounded so relaxed. My only complaint is that, though she takes all repeats, she doesn't vary the ornaments.
Evgeni Koroliov's Goldbergs (Hänssler) features liberal but tastefully applied added ornamentation and all variation repeats. It's not perfect: in Variation 12 in the repeat of the first half, there's an awkward hesitation in measure 5; in Variation 18, he plays repeats an octave higher, an effect designed to simulate the registration change of a harpsichord that's too crude on piano (Andras Schiff's overuse of this device, suggesting a music box rather than a piano, ruins his recording). But Koroliov's technique is impeccable, with trills very even, cross-hand passages handled smoothly (even in the tangle-inducing Variation 11), and excellent balance in Variation 28 between the buzzing 32nd notes and the outer parts. Though there is legato in appropriate spots, he favors a detached, nearly marcato style, which provides great clarity yet is not too dry in effect since there's a natural amount of resonance in the acoustic (this recording boasts superb sound). Most of all, there's a stimulating marriage of intelligence and spontaneity, not only in ornamentation but also in unmarked arpeggiation, impeccable rubato in Variation 6, occasional flexible tempos, and a transformed Aria at the close.
There are a couple Goldberg icons I'm not impressed by. Pianist Rosalyn Tureck made five recordings of them, with increasing pedanticism producing correspondingly stultifyingly results. And in both her 1933 (EMI) and 1946 (RCA) recordings, Wanda Landowska's clangorous modern Pleyel harpsichord makes unbearable sounds that are anything but authentic. Musically, there are also objections (that go far beyond the lack of repeats). She's sometimes rather Romantic in her phrasing, and her registration changes (for instance, in Variation 7) can be coy to the point of preciousness. Her tempos in canons can sometimes be sluggish as she attempts to clarify the lines in spite of her instrument. She even alters the music at a few points, changing the endings of three Variations.
Of course, our knowledge of genuine Baroque harpsichords has vastly increased in the intervening decades, and given that Bach wrote this masterpiece for that instrument, it should be heard on it as well as on piano. First choice is the 1992 recording by Pierre Hantaï on Opus 111 (his more recent remake is a bit too mannered to be a top recommendation, though it's fascinating in its own way). It does have its peculiarities.Hantaï "splits hands" in the Aria (the left-hand notes sounding fractionally but noticeably sooner than the melody), though only rarely does this recur in individual variations. He takes all repeats except in the melancholy Variation 25 (and also inserts some unnotated arpeggiation), one of the variations where Bach actually wrote in first and second endings; perhaps Hantaï feels that the emotional intensity of this slow movement could not stand up to repetition. On the issue of ornamentation, even the first time through some sections he adds ornaments beyond what's noted, and often adds more the second time through, for instance in the second section of Variation 18. He uses some expressive hesitations, but neither too frequent nor too drawn out, except perhaps in Variation 27. He takes Variations 5 and 8 at a relatively slow tempos, which allows him (especially in 8) to insert rubato. Oddly, he rushes at times, for instance in the last two bars of Variation 3 (both times). Like a few other harpsichordists, he plays Variation 7 with a duple rather than triple feel. Variation 16, the Ouverture in the French style, is played with fabulously idiomatic verve. A major attraction of his recording is the harpsichord itself, a 1985 Bruce Kennedy copy after a Michael Mietke instrument from the very early 1700s. Its silvery treble and rounded tone throughout its range hint more at a clavichord sound than most harpsichords, and make it less clangy. The 32nd notes in bars 13-16 and 29-32 of Variation 14 flow with a wonderful liquid tone on his instrument.
Also worth hearing is Blandine Verlet on Astrée/Auvidis. She takes all repeats and varies her ornamentation the second time through, takes a generally lively but not exaggerated view of tempos, handles registration well, and uses agogic pauses tastefully but not excessively. Anyone who finds Hantaï too mannered should be quite satisfied with Verlet instead. Extremely economy-minded harpsichord lovers will appreciate Maggie Cole's two-CD set on Virgin Veritas, with a stylish Goldberg Variations on disc 1 and the Chromatic Fantasia & Fugue, Italian Concerto, Partita No. 1, and more on disc 2, all for the cost of a single disc.
When Johann Gottlieb Goldberg died on April 13, 1756, he had not yet reached his 30th birthday. Now there are well over 200 recordings of the piece that bears his name. - Steve Holtje
Mr. Holtje is a Brooklyn-based poet and composer who recently recorded his original soundtrack to Bystander, a documentary film by John Reilly.