One afternoon in 1920, a young pianist sat down in a shuttered room in the capital of defeated Germany and played a Bagatelle by Beethoven. At the return of the main theme, one of his fingers fractionally strayed, touching two keys instead of one. 'Donnerwetter!' (dammit!), cried Wilhelm Kempff. He looked around and saw crestfallen faces. 'That was very beautiful,' said the machine operator, 'but the recording is now ruined.'
This lapse, recalled by Kemp years later, amounts to a defining moment in the annals of performance - the moment a musician realized that recording required a different discipline and temperament from public concerts. Kempff, had his finger slipped on stage, would have played on regardless, knowing that few would detect the fiaw, or remember it afterwards. On record, though, the imperfection was engraved for all time, growing larger and uglier with each replay. There was no hiding place, no camoufiage available on disc for inferior technique or inchoate interpretation. The artist stood exposed to eternal scrutiny, stripped of illusory diversion.
Sound recording had begun in 1877 with the inventor Thomas Alva Edison shouting 'Mary had a little lamb' into a phonograph and acquired a mass market in 1902 with the first brass-horn arias of the Neapolitan tenor Enrico Caruso. But the birth of recording as a musical act, separate and distinct from live performance, came in 1920 with the undeletable exclamation of a German artist in the aftermath of the First World War. Kempff, a protege of Brahms' friend Joseph Joachim, was rooted in gaslight romanticism but suffciently aware of swirling currents to realize that recording presented more than just an opportunity to earn a fee. What it offered, once an artist had overcome the fear of error, was the chance to achieve a perfect score. For the first time in cultural history, accuracy and speed transcended inspiration as the object of performance, and there was no shortage of young men like Kempff who wanted, quite literally, to set a record with their playing.
Wiser heads demurred. The professional pianist Artur Schnabel, a man of lofty mind and caustic wit, argued that recording went 'against the very nature of performance' by eliminating contact between player and listener, dehumanizing the art. Music, he said, was a one-time thing, once played never to sound the same again. Schnabel turned his back monumentally on mechanical impertinences. Kempff, meanwhile, faced fresh dilemmas, moral and aesthetic. Recording, he discovered, was innately competitive. Where, before the war, no one could have asserted empirically that Ferrucio Busoni was a better pianist than Ignacy Jan Paderewski, now it was possible to measure Kempff against Wilhelm Backhaus and, music in lap and stopwatch in hand, checking every note in the Moonlight Sonata and timing each movement against Beethoven's metronome mark, prove that Kempff was materially superior. Strife ensued. Artists became bitter enemies and listeners were confused. Soon, it was not enough to have one Moonlight in the living-room cabinet; two or three sets displayed intellectual breadth and civilized tolerance. Where emperors in Vienna once staged live contests between Mozart and Clementi, the suburban homeowner in Peck-ham or Pittsburgh now played Rachmaninov against Vladimir Horowitz for a satisfyingly close shave. An element of sporting competition entered the musical game.
Kempff, who lived to the great age of ninety-five, was a studio master. His articulation was explicit, the notes separated as if bejewelled, his interpretations eschewing an excess of individuality. He recorded the popular classics twice, bought a castle near Bayreuth and was exclusive to Deutsche Grammophon from 1935 to his death in 1991. Yet, while his records entered thousands of homes, Kempff was never a household name. Lacking stage magnetism, he did not visit London or New York until 1951 and many who queued for hours to hear Kempff repeat his estimable studio interpretations came away feeling defrauded. Where was the raptness, the subtle variants of colour, when this nondescript little fellow sat upon an empty platform? Kempff, they complained, was a synthetic invention - a soloist who could never have flourished before the anonymity of recording. His fame came from work done in the dark, away from social and political realities. In his memoirs Kempff is untouched by the century's traumas, by Hitler or mass hysteria, unaware that, when he played in occupied Krakow, he was less than an hour's drive away from Auschwitz.
Schnabel, by contrast, was acutely attuned to public mood and eventually dropped his resistance to recording on an assurance that his work would be sold only in Europe and the British Empire until American audiences had a chance to compare his living presence with the shellac substitute. The principle of eye contact remained uppermost in his mind. Gregarious and polyglot, a commanding presence at the keyboard, Schnabel created a new edition of the thirty-two Beethoven sonatas and played them serially, start to finish, in seven Berlin recitals for the 1927 centenary of the composer's death. He repeated the cycle twice in London while recording for His Master's Voice. The last box in the 100-disc series, sold by advance subscription, appeared in 1939. Schnabel, in this set, introduced a twin-edged concept of integrity: the complete works, performed by the supreme authority. But the idea of the complete cycle had another advantage in that it sold people things they never wanted or knew existed. Subscribers who signed up for the Moonlight, the Hammerklavier and the imposing opus 111 received, together with these summits, discs of less interesting sonatas. Schnabel's Beethoven showed that great composers could be marketed to the self-improving middle classes as a mantelpiece essential, like Encyclopaedia Britannica, the plays of Shakespeare and a potted aspidistra.
Schnabel did not take easily to recording and the producer had to bring in his pretty niece to turn pages to give him an illusion of audience. 'I suffered agonies and was in a state of despair,' he reported. 'Everything was artificial - the light, the air, the sound- and it took me quite a long time to get the company to adjust some of their equipment to music.' The recordings, however, are the antithesis of synthetic. They ripple with spontaneity and are riddled with wrong notes, scintillating in their contempt for precision and their search for inner meaning. Schnabel, said the Chilean pianist Claudio Arrau on his death in 1951, was the first 'to illustrate the concept of the interpreter as the servant of music rather than the exploiter of it'.
His record allies had no qualms about exploitation. They took Schnabel's notion of integrity and sold it as doorstoppers to a world that furnished its homes with big boxes. If Kempff 's expletive defined music ex machina, Schnabel's blessing put the whole of Beethoven within mundane domestic reach.
Sounds that were collected before these events are chiefiy of archaeological interest. To listen through aural debris to Francesco Tamagno (1850-1905), Verdi's original Otello, or to Alessandro Moreschi (1858-1922), the last castrato, is a fascinating experience but one that cannot be endured for much longer than holding one's head down a wishing well. The pitch is wobbly, the static obtrusive and any impression of the singer's musicality requires an imaginative leap on the listener's part. Mighty Melba comes forth enfeebled, Tetrazzini underpowered, Galli-Curci unbeautiful. Mint copies of these objects fetch thousands of dollars (a prolific collector was the oil billionaire, John Paul Getty), but artistic satisfaction is hard to come by on these hand-cranked receptacles.
The first recordings to overcome extraneous noise were ten arias taken by a young American, Fred...