Written by Colin Tilney

Then and Now: Learning to Write for the Harpsichord

Louis Couperin and Dieterich Buxtehude had little or nothing to learn, of course. If you were French in seventeenth-century Paris, you wrote dance suites for the harpsichord; if German in eighteenth-century Lübeck, you sat down at it and turned a comic popular song about the wrong kind of food into a set of brilliant variations. The instrument was central to musical life, the piano of Then, perfect for displaying fingerwork in solo recital and, more important, indispensable for the rehearsal and performance of opera, oratorio and orchestral works. Some composers naturally made greater imaginative contributions to the repertory than others: the uncle of François Couperin, for instance, had borrowed the concept of indefinite note-length from the lute and used it for the notation of his mysterious unmeasured preludes; Bach found an hour’s music in a simple sarabande. But in general there was no need to fret about writing idiomatically for the harpsichord. You played the instrument every day and that was enough to be convincing.

What about Now, though, or more precisely the period just within living memory? By 1800 the early piano had effectively dethroned the harpsichord, which remained silent for a century or so. During that long time some French and German amateurs occasionally dressed in period clothes and played recorders and lutes together for a few hours; in England at the end of the century Arnold Dolmetsch did that too, but professionally – and when he built instruments, he copied antiques as accurately as possible, in his search to rediscover the music of the past, exactly as it had been heard then. That movement, later called the Early Music Revival, grew slowly through two world wars to become in the fifties a vital factor in entertainment and education, with the founding of specialized schools, widespread building of historical instruments following traditional methods and the training of performers in past  technique and theory. How well did the harpsichord do in this new dispensation? Very well indeed, both in the variety of models available and in the acceptance of the player as both soloist and accompanist. Continuo was transferred from the harmony exam to the concert hall, and every string band now had its own harpsichordist, in accordance with C.P.E.Bach’s dictum that “no piece can be well performed without some form of keyboard accompaniment.” (However, a brief episode from an Edinburgh Festival rehearsal of Vivaldi string concertos will show that not everyone shared Emanuel’s views. Great Russian cellist: “Harpsichord too loud – pianissimo, less, nothing. Don’t play!”).

Like Lazarus, the harpsichord had risen again – and everyone started writing for it, in astonishing numbers. The magisterial Harpsichord and Clavichord Music of the Twentieth Century by Frances Bedford covers all solo, chamber and orchestral music for early keyboard written between 1900 and 1992: she lists 2,304 composers. Much of the work is of academic origin from the many American universities offering degrees in music, but some mainstream names turn up: Bartók (OK to adapt Mikrokosmos), Busoni, John Cage, Falla, Krenek, Ligeti, Milhaud, Arvo Pärt, Poulenc, Schnittke, Stockhausen (harpsichord as admissible candidate), Richard Strauss (opera arrangements), Xenakis, as well as many more who are only slightly less famous nowadays. Debussy would have included harpsichord in his fourth (unwritten) sonata. Sadly missing from the serious solo section are Stravinsky, Britten and Michael Tippett, all of whom included the harpsichord as acoustic flavour in some of their operas but might have used its sonorities to devastating effect at a deeper emotional level. (In only one short scene in The Rake’s Progress, for example, is the harpsichord excused eighteenth-century pastiche.) With so much to imitate and no help from the great break in tradition, the challenge for composers was to be inventive, but at the same time efficient. What would work on the harpsichord and what wouldn’t? One of the major mistakes to avoid was aping the piano. “Dance for Harpsichord”, the lovely foolish gavotte that Delius wrote for the English pioneer, Violet Gordon Woodhouse, is unplayable on the harpsichord – for want of a sustaining pedal! Copying the eighteenth century provided Stravinsky with some of his best ideas, but it is noticeable that he never uses the harpsichord in his would-be Baroque scores, and there has probably been far too much neo-Classical charm offloaded on to the poor instrument, giving it sometimes an undeserved air of frivolity. Certain acoustic considerations are relevant, too. Complicated multiple strands of counterpoint are a trap, particularly if there are more than two and they are rhythmically confusing. (Some of Henze’s harpsichord writing fits this category.) On the other hand the harpsichord has much to attract a composer: bright attack and intriguing decay; clear vocal registers; sharp contrast between notes held and notes released; above all, it delivers compelling rhythm. And its natural  attraction to conventional harmony can at times be fruitfully disguised as bi-tonality, where each key adds spice to the other but both remain friendly (listen to the supreme example of this in the slow movement of Falla’s masterly  concerto). Unrelieved dissonance, as in some serial writing, does the harpsichord no favour.

All four twentieth-century composers in tonight’s program seem to me to enjoy the harpsichord and treat it well, although they approach from different places. The dissonance in Ohana’s Carillons is accentual, not serial, and totally joyous, representing by clusters and out of phase scales the sounds that keep the happy traveller in the Low Countries awake at night. Here the major sevenths and minor ninths are atmospheric, exhilarating but not harsh, with much brilliance from the octave strings. By contrast, Andriessen uses throughout only the two 8′ registers, one on each keyboard. Commissioned to write the piece for a Dutch harpsichordist, Louis resolved to deny entry to all pianists, perhaps to reprove them for their part-takeover of the Goldberg variations. In Overture to Orpheus every note on the lower manual is repeated instantly on the upper. A young and very energetic pianist might possibly be able to mimic the sequence, but not the difference of colour involved, the very deception that is forced on pianists in Bach’s two-voice variations. Japanese koto playing lies behind the sound of this music.

Next come two commissions. Priaulx Rainier was born in South Africa, but lived in England from 1920 until her death. I gave the first London performance of Quinque at the Purcell Room in March 1974. Priaulx was a professional violinist, but she had amazing insight into harpsichord sonority: “In writing this piece I have kept in consideration the richness and at the same time the limitations of the instrument.” She understood perfectly the need for some notes to be long, some quite short, and she writes in bursts of sound, gestures, that border tonality – but always in surprising ways. The five movements are fast, slow, faster, slow (a monody) and very fast. The book says it is “difficult”. It is stretching, wonderful music – ideally written for the harpsichord, though not yet published – and I feel I can hear in it the bird and animal sounds of her African farm. Bedford does not mention Linda Smith’s Gravity, written for me in 1988, although it dates from just slightly ahead of her cut-off point. Linda writes: “The title refers to the idea of the low notes creating an illusion of gravity, and also to the way pitches collect themselves, or gravitate, into harmonic worlds. It is written with an allusion to the unmeasured preludes of Louis Couperin, where time can be interpreted quite freely.” This is a deeply satisfying journey through harpsichord sonority, full of sumptuous harmonies that are rarely entrusted to the instrument.

More about the program. The “free interpretation” of Couperin’s prelude is my rhythmic response to a notation consisting entirely of whole notes, enhanced by  decorative lines that appear random but in fact give vital information about note-length. The movement has three parts, its central section being a miniature fugue in conventional notation. The following dances in the suite observe the forms of the time, although Couperin’s thought and his harmony are far from ordinary. The melody in the Buxtehude variations comes from one of the two folksongs in the Goldberg Quodlibet, Kraut und Rüben haben mich vertrieben  (“Cabbage and turnips turned me off; if only my mother had cooked us some meat, I might then have stayed here  longer”.) As composers, both Buxtehude and Louis Couperin probably offer a more novel and engaging contrast to the others on the program than Bach, Handel and Scarlatti would have done; they are certainly less likely ever to be played by pianists.

More still. I am aware that not one of the “modern” pieces I have chosen was written in this century, just as none is scored for prepared harpsichord, requires electronic equipment or is paired with exotic ethnic partners. We might call the program a kind of retrospective, mixed with respectful attention to some of the things going on today; we might say it was a bit old-fashioned, perhaps. I have tried to explain, above, what I believe the harpsichord does best and how the instrument resonates more richly the nearer it comes to playing a simple triad. It isn’t a piano. I think we shall have to leave it there, headed to an unknown future. I’d bet on its survival.

Many thanks to Jim Anderson, my dompteur de clavecin.