I think it was Morton Feldman who said, famously, that ‘a composer that doesn’t have a friend who is an artist is in trouble.’ His friends were among the abstract expressionists providing the visual wallscape of his New York loft: Robert Rauschenberg and Sonia Sekula. For Harrison Birtwistle his special ‘friends’ amongst artists were painters he never met, but their work has held a lifelong fascination. Paul Cézanne and Paul Klee have been the dominant influence, and it’s not difficult to find parallels between their preoccupations and his. I ‘m told you are more likely to find Birtwistle looking at pictures than listening to music, the latter something he declines to do except when he’s preparing an opera and would ask his late wife Sheila to take him out in the car – with the stereo playing Verdi.
My association with Birtwistle has been since my teenage years when I heard the first concerts of the Pierrot Players, the ensemble he formed from his colleagues and associates at what is now the Royal Northern College of Music. It seems as he has always been a presence in my view of contemporary music, and remains so. In his eighties now he is still producing fine work, from his recent piano concerto titled Responses to the beautiful miniature for guitar I heard just last week – and which I’m playing in a recital next year.
One of the reasons I continue to write this fortnightly blog is that I’m insatiably curious about other composers’ music. Thirty years ago I wrote an essay that attempted to answer the question How Composers Think. It was this essay that began my association with script-based algorithmic music. It seemed a question worth asking because in order to use computer-aided techniques I had to try to understand the way I thought. This process of understanding your own way of getting those notes on paper and the choice of materials and techniques you use seems an essential prerequisite of composing successfully. I’ve focused on this during the early chapters of my book Parametric Composition.
Whilst Birtwistle is notoriously reticent about talking of how he composes, he has left his voluminous sketches and drafts to the Paul Sacher Stiftung. For any scholar this formidable library provides the key to how Birtwistle’s music is made. I know myself how revealing and extensive the body of work that goes into a major composition can be. I have two filing cabinets of folders containing such material that go back at least thirty years – sadly (perhaps) the material before 1986 has been ‘lost’ or destroyed. From the time I started composing with computer aided techniques I’ve also favoured the score-file annotation, and readers of my book will know these accompany the full-scores of my work and can be downloaded from my website.
Whilst such information can be revelatory I enjoy the challenge of working things out myself, and this is what I’ve been doing recently with a Birtwistle score Refrains and Choruses. If Birtwistle went in for opus numbers this would probably be his Opus 1. I’ve chosen it because I’m composing a large-scale wind quintet and I’ve been studying the way different composers address the medium. For Birtwistle the medium of wind instruments is particularly present in his work. He studied the clarinet, and it was only the success of Refrains and Choruses that was the signal to sell his instruments and be a full-time composer. It is also well known that amongst his primary influences, i.e. scores that he studied meticulously (do people still do this today? – I get the message they don’t – maybe, as Brian Ferneyhough is wont to suggest, it is the CD that’s important, that is the text). For Birtwistle, Stravinsly’s Symphonies of Wind Instruments and Octandre by Varese exerted a strong influence and fascination. Later it would be Messiaen’s Et exspecto resurrectionem mortuorum and Stockhausen’s ZEITMAßE. The latter work was written in the same year as Birtwistle’s first essay into the wind quintet. All these aforementioned works are for wind instruments.
The Stravinsky score put a particular spell on Birtwistle, as it has done for me. Curiously enough, it was the amazing solo piano version of the Symphonies that seemed finally to reveal its true nature.
What is it about the wind instrument and wind ensemble? Is it the nature of breathing that we all own. String instruments for me are all about gesture and bodily movement. Wind instruments are closer to ancient ritual, the Greek aulos, that pipe of Marsyas defeated by Apollo’s Lyre. In my own output wind instruments seem increasingly a medium of choice. My first major piece is EDGE, a work for wind quartet written with algorithmic means but just prior to using script based languages.
My own particularly favourite combines piano with a wind quartet, Mozart’s beloved combination without the flute found in in his Eb Quintet. My own Quintet takes the Mozart very much as a starting point.
In Birtwistle’s hands the wind instrument has something earthy, mystical, elemental about it. It is these characteristics that so often are embedded in his music.
Sadly Refrains and Choruses has recently been removed from Youtube, but there is a good recording which you can access and buy separately as an MP3, although the CD it comes from is excellent and contains a much later work for wind quintet titled Five Distances.
In this blog I want to focus on the last 32 bars of Refrains and Choruses (this link takes you to an on-line PDF of the study score). This is probably the most direct music of the 9 minutes piece and one that sets up an intriguing problem for the script-based algorithmic composer. The ‘refrain’ here is a single chord, a pitch class set (cs4 d4 e4 a4 bb4) which plays eight times between ‘choruses’ of duets using various different combinations of the quintet. Whilst the choruses show a rare example of Birtwistle in (almost) serial mode the punctuating chords throw up an intriguing compositional problem. Birtwistle begins with the chord in its widest manifestation (d2 a3 cs4 bb4 e5) and gradually compresses it chord by chord to (a3 bb3 cs4 d4 e4). Now, how to do such a procedure algorithmically? It’s a real test, and one I thought I would try to surmount. If this ‘problem’ was for a keyboard scoring I can see a way, but to have to consider and keep within the ranges of each of the wind instruments that’s a different matter! What I’m gradually learning about constraint programming begins to show a way, but this is not yet available in the two software applications I use. And anyway, for the composer working on paper such a process is an easy 5-minute task of visual scanning – as I know the practical ranges of the wind instruments and mapping a serial row sequentially across duets is no big deal, rhythmic material aside. The half hour I spent playing with this ‘problem’ convinced me that it’s all too easy to think script-based programming should produce such relatively straightforward procedures, but the bitter truth (for me at least) is that there are aspects of programming we can hardly begin to deal with. I’ll append my partial solution to this text. I’ve annotated the script a little to show how some of the procedures I’ve used could well be valuable in other contexts. I shall be intrigued to see if any coders out there can produce a workable solution. Something to occupy the mind over the Christmas Holidays perhaps?
If you do manage to see the whole score there’s another device to observe that has become very common in Birtwistle’s music. It has become known s the ‘wedge’. This has its origins in the writing of Paul Klee’s, in his Pedagogical Sketchbook, a book Birtwistle has often referred to a major influence on his thought and procedures. We all know Klee’s ‘taking a line for a walk’, but the ‘wedge’ is relatively unknown.
This image comes from a valuable introduction to this Sketchbook to be found at the Tate Gallery website. It is all about working in two dimensions and the idea and potential of the wedge as a mode of thought is neatly shown in Klee’s diagrams. Klee was a musician, a skilled violinist, and wrote quite extensively about music. Here he is about to play the Schubert Quintet in C.
In Refrains and Choruses the passage from 89 to bar 122 shows how the wedge, secured around a pedal E (Birtwistle’s favourite ‘starting a piece’ note, generates an increasingly complex passage. You can read about this device in detail in a discussion of my score Serenade in Chapter 16 of Composing: How and Why (the on-line version of Parametric Composition. Serenade and Trio are two pieces from A Selection of Sextets published to coincide with the launch of Opusmodus. Both scores use the gen-curve algorithm, which gets you some way towards a wedge-like process.
Coda: whilst writing about Birtwistle it seems a good opportunity to discuss his long-standing use of random numbers. This is pretty much a pre-requisite of computer-aided composition. Randomization takes so many forms, either as a way of varying existing material (processing) or actual generating new parametric elements. We know from Birtwistle’s manuscripts in the Paul Sacher Archive the composer began using randomization around 1960. In his ‘Book of Magic’ compiled from about 1967, and in use until 2003, we can see list of numbers pencilled in drafts of scores, a tie in with typewritten sheets of random numbers Birtwistle commissioned from a colleague about the time of writing Refrains and Choruses. Although there’s no evidence they were used in this wind quintet there is plenty of evidence of their use in later works. There’s a reported instance in his Carmen Arcadiae Mechanicae Perpetuum where he assigns integers to specific dynamcis pp (1,2,3,4,), mp (5,6,7), f (8,9,), fff ‘(10).
Birtwistle’s use of these numbers has made me stop and reflect about randomization and the way I occasionally use it. His ‘method’, basic in the extreme, allows for a measure of reflection that computation doesn’t always offer. It’s just too easy to use randomization in a very large-scale fashion, whereas it is clear Birtwistle often uses only tiny fragments as source material, particularly for textures in orchestration. In an interview, when he mentions ‘the numbers (10-digit number series – but over varying lengths between 1 and 10) I keep in a draw of my desk’, he explained their use in a flurry of woodwind tutti in Earth Dances. So in Birtwistle’s way of things a very few numbers may go a long way, as his brain computer has simply become used to making instant transformations from integers into pitches, dynamics, maybe even rhythms., and just in the same way he organised his eight chords in Refrains and Choruses. We don’t know, but clearly a little goes a long way!