Piano Improvisation by Sir Edward Elgar’s
This post was published by Norris and asserts his opinions.
On Nov 6th, 1929, in the Small Queen’s Hall in Central London, Sir Edward Elgar improvised at the piano straight onto HMV’s wax. The resulting five pieces were never heard live again until 1995, when I gave their second performance ever, playing by ear, not from notation, in the Bergen Festival. I have played them frequently since then.
Elgar was used to recording as a conductor, but this was something new and strange for him. To settle himself down, he began with a meditation on a ballet tune from Rossini’s William Tell. It was Elgar’s life-long habit to improvise his way from the known to the unknown – that’s how the Enigma Variations came into being – and his delight in turning a typically perky Rossini melody into a characteristically yearning bit of Elgar got him off to a good start with his new project.
There followed two pairs of pieces, in G and in D (though ending in A flat). A searching introduction leads to a broad melodic snippet from The Fringes of the Fleet, treated with ever more imposing flurries of arpeggios and octaves. Then a Rondo, started and punctuated by 4 enigmatic chords. (The Rossini dance begins with slow chords, which may have caught Elgar’s fancy.) The second pair of Improvisations begins with a ternary-form Intermezzo which is essentially an enlargement of the slow movement for piano and orchestra which Elgar apparently gave to Harriet Cohen in short score, and which was orchestrated by Percy Young.
Dr. Young has kindly shown me his version, but I have not seen the Cohen manuscript. It presumably pre-dates the Improvisation, which adds a new central section. Finally, a darkly majestic opening in D minor introduces a beautiful two-section melody in C major. Some strenuous harmonic excursions lead Elgar to A flat, where he ends rather suddenly, probably because Fred Gaisberg was making winding-up gestures to let him know he’d got to the end of the side.
This neatly introduces the dilemma of one who would re-create an Improvisation. I am not constrained by the brevity of a 78 side. Should I stop where Elgar had to? That last Improvisation follows an intricate pattern of thought, with quasi-recapitulations and fleeting thematic references. I’m convinced that Elgar would have wanted to recall his beautiful melody. Its opening falling sequences, static in harmony, lend themselves to circular imitation, which is very effective as a close. The second limb of the melody, with rising harmonies under its falling phrases, I present as a final climax in combination with the idea of the introduction. Then I play Elgar’s final cadence, with its unmistakable reference to the word “wiedersehen” in the soprano aria in the Brahms Requiem.
Elgar improvised all his life for himself and his friends, to share ideas and to experiment with structures to contain them. His Improvisations were not intended to be polished pieces, with every harmonic problem stitched up, and all the notes under the fingers. And so, in these recorded Improvisations, there are several moments of utter incoherence. Rather than reproduce them exactly, I have plotted a more considered course round the same harmonic rocks and melodic rapids.
Apart from these two sorts of tidying up, I began work on learning these Improvisations by finding the very notes Elgar played. Michael Mullen of the BBC Music Library gave me invaluable early assistance by sketching a score of the structures, but the detail was a matter of months of listening. There were various classes of detail, beyond getting the tunes and the bass and the harmony. The exact spacing of a chord could make a huge difference to the emotional resolution of a passage.
Some of the details were details of absence, dogs that didn’t bark, as it were. I would labor over some tricky modulation, using all the resources of chromatic harmony to try to clarify it, when a rehearing would reveal that Elgar had turned the awkward harmonic corner by using only his characteristically ambiguous three-part harmony. The tonal difficulties melted away in the face of Elgar’s resolute determination not to deal with them.
One important class of detail was an absence of detail! Elgar was very proud of his ability to play the piano with an apparent abundance of orchestral effect, the despair of conventional pianists. This orchestral jangle was produced by busy “faking” of runs and arpeggios, and is heard throughout the G major Rondo. When one tames these flourishes into something that could be written down, both as to notes and rhythm, they lose all their effect.
One is reminded of the time Richard Strauss conducted an orchestra that could actually play the opening of Don Juan. The thrilling, flying string scales came out rat-a-tat-tat, and Strauss threw down his baton because he didn’t want it to sound like that! So in the Rondo, I recall my days as a repetiteur at Covent Garden, and I scramble and tremble and thunder.
This feeling of freedom from notation turns out to be central to the Improvisations. When we came to record them, I began by taking care to play what Elgar had played in 1929, as exactly as I could, playing from a score in my head, as it were. Then we recorded the pieces as they have formed under my hands in frequent performances over the last five years. This time it was not a score in my head, but a sound. In every case, in the opinion of the entire team, this freer version came closer to the music.
This reinforces work I did while I was the Gresham Professor of Music, investigating the effect of notation on performance. It’s most succinctly summed up by the story of an Indian musician, unfamiliar with Western notation, who listened to a phrase of 16th century music on a recorder, played first from the original notation, with its odd note shapes and funny clefs, and then played from a modern transcription. The performances were identical to Western ears, but the Indian musician could tell them apart, and got them right every time.
The act of Notation does something mysterious to music, something which became part of the concept of Classical Music, which in fact I define as “music that has to be written down”. In passing, I might note that this crucial aspect of Western classical music is being lost, as more music students learn music from recordings before they open the score.
My recording of the Improvisations, to be released on Nimbus, sheds a number of unusual lights on the relation between score and sound. The Improvisations show Elgar first working from another composer’s score, then working on ideas he had himself written down (though in the D major Intermezzo he scarcely plays the tune the same twice!), and finally genuinely launched on an unknown sea of spontaneous creation. Yet the whole point of the project seems to have been to help him with the composition of his inevitably written-down Piano Concerto. And in Robert Walker’s work, we can hear the fixing process of notation at work in the material he has drawn from the Improvisations.
I played these pieces recently on the very piano at which Elgar improvised his way to the Enigma variations. It was an uncanny feeling of identification with him. As I follow in his musical finger-steps, may my performances share some small part of that blessing with other lovers of the great composer.