Several useful points for composers to consider came out of our recent workshop of new music for two harpsichords:
- Who wants it?
There is no significant professional market for new music for historical instruments. No-one (currently) makes a living from principally performing new pieces full time.
In the professional context, a new work usually appears as an added extra – perhaps as a commission inserted into an otherwise baroque programme. This immediately reduces the ongoing scope for any new work requiring a special set-up, tuning or equipment beyond that available for a baroque recital.
Very little new music for two harpsichords has made it into the repertoire, and even the solo contemporary repertoire (more than ten thousand pieces) contains few that are well known by players (the three virtuoso works by Ligeti are a good example, on grounds of their outstanding quality).
- First impressions count
Avoid putting the performer off by a work’s presentation before they’ve had a chance to form an opinion on the music itself.
- Scores should normally be in portrait orientation. Landscape scores cause problems in filing and binding, especially if performers are playing from print-outs in a ring-binder (which they often are).
- Use 6mm minimum staff size, and be aware of page turning issues.
- Enharmonics: accidentals must be expressed in the most straightforwardly understandable way possible. Unnecessary barriers to understanding may put performers off. Don’t only rely upon the Sibelius plug-in to make the most appropriate enharmonic choices.
- A score must give the right instructions for a human performer to achieve the desired result – avoid littering it with markings which appear solely to facilitate Sibelius playback.
- Clarity is critical. If a performer has to spend too much time deciphering unnecessary obscurity on a first reading, they may not bother with a second one. See also Dylan Christopher’s useful article on this topic.
- Consider the performer
Bear in mind the comfort zone of harpsichordists. The harpsichord is not the piano, and someone who has chosen to play it has by definition not set out to reproduce pianistic effects and techniques. One particular bugbear is left-hand octaves – common on the modern piano, but less comfortable (although sometimes quite effective) for a harpsichordist.
It’s worth remembering that the usual baroque repertoire of harpsichordists tends to a limited range of key signatures. Anything beyond three-four sharps/flats will be rather less familiar for a harpsichordist than, for example, a pianist or organist.
- Consider the instrument
Perhaps fewer than a quarter of existing harpsichords are double manual, and those instruments tend to live in institutions, concert venues, or are owned by baroque professionals. Composing a piece specifically for two-manual instruments therefore cuts out most of the amateur market.
For compositional purposes it is best to assume a single manual, GG-d3 compass (see wikipedia article for info on ranges) with two 8′ stops and buff. These are widely available, and access to a full-size professional five-octave double manual harpsichord should be regarded as a bonus not a norm.
Real instruments vary so much that it is usually best not to specify particular stops. Give instead a dynamic marking, or a mood or tone colour which the performer can then implement appropriately with what is available, given the voicing, tone, acoustics etc.
- Consider the context
Why two instruments – what is added that can’t be done with one? Is it a dialogue between two voices, a soloist-accompanist scenario, or perhaps a competition between the two players? Does it allow for a particular expanded formal structure or certain types of repetition-dialogue?
Also, what is the piece for? Is it a virtuoso set-piece to sit within a public Bach recital, or a ‘social’ piece for harpsichord-owning friends to play in private? Is is an encore utilising the somewhat-neglected humorous aspect of the harpsichord, or an experimental ‘paper piece’ designed for coursework submission?
These all feed into rehearsal time constraints. If, for example, a complex piece will take as long to learn as several simpler works, it’s unlikely to be feasible for a recital involving a lot of new material. It would be far more likely to appear as a single new item in a programme of known repertoire.