I am an experimental composer

I presented the following paper at the COMPOSITION‐EXPERIMENT‐TRADITION conference at the Orpheus Institute in Ghent in February 2012. Because the text is largely speculative, informal, and anecdotal, it isn’t particularly appropriate for a more scholarly publication, but it is one that I nevertheless wanted to make publicly available, so I am posting it here. The middle section of the paper originally included some extemporaneous references to examples from two pieces, Being itself a catastrophe, the diagram must not create a catastrophe (2007-09) and Second String Quartet (2010). I haven’t attempted to recreate those examples here, largely because those materials are now included in publications for Sonic Ideas and Contemporary Music Review, respectively.


Orpheus Research Centre in Music [ORCiM]
22‐23 February 2012
Ghent, Belgium

I am an experimental composer

(or, ‘How do you know if you know what you’re doing?’)

Many years ago, in the earliest iterations of my composer bio on my website, I described myself in the first sentence as ‘an experimental composer’.  A German friend & colleague persistently made fun of me for using this term – I was never quite sure why, but in any case I eventually deleted the offending line from the bio.  It only began to make sense last year during a ‘bloggingheads’-style interview that I did with the composer Michael Pisaro.  He retold a story of a conversation he had with the Austrian composer Peter Ablinger, in which Peter asked that Michael not describe him as an ‘experimental composer’, because in the German-speaking world that term implied that, as a composer, you didn’t know what you were doing.  (He did add, parenthetically, that it would be fine to call him ‘an experimental composer in the American sense’.)

Despite my willingness to remove the term, the fact has remained that what I do as a composer – and why I do it – is driven by a dedicated prioritization of experimentation.  Virtually every stage of the compositional process, and indeed the compositions themselves and their performances, revolve around an effort to foreground a practice of experimentation – that is, a process that is driven by questions, rather than answers, hypotheses, rather than conclusions.  Most critically, what has been central to my work is the fact that the work itself is the experiment, rather than what seems to me to be a more common practice of exploring/experimenting before the work, with the composition appearing as a statement or answer or at least provisional conclusion.

I’ll look very briefly at three areas of my work:

  1. experiment in compositional practice (experimentation through discovery, emergence, investigation; that is, ‘the work as experiment’)
  2. experiment in performance (experimentation through unpredictability, instability, and indeterminacy)
  3. issues and questions – both artistic and ethical – that emerge from this approach

To begin, it seems necessary to outline a few basic principles of experimentation:

  1. For me, experimental art is principally defined by its effort to prioritize questions over answers – those questions tend to be driven by investigations into what might be possible and what the implications of that possibility might be; it’s crucial, by the way, that the answers to these questions are fundamentally unknown (that is, simple, practical questions to which I don’t yet have the answers aren’t sufficient)
  2. An effort to explore uncharted territory – that is, at least at some level, a mapping of the unknown … (this is why, for example, I make a distinction between ‘an experimental composer’ and ‘a composer of experimental music’ – the latter for me could include composers of a particular style(/brand) of music within a terrain that is already mapped, whereas the former is principally about the artistic/creative process)
  3. A certain (if sometimes quite loose) connection to the scientific method.  More informally, the issue is that the pieces are seen as a kind of testing ground for a particular question, but, beyond that, there is a need for a clear path of resistance, one that defines the space of the experiment.  Open-ended experiments are not experiments as such, as what is required is a closed, constrained, limited space.  Experiments need to in some way push back against or provide resistance to their subjects.
  4. Following from this, a focus on systems and a certain precision in the construction of a work, or, more importantly, carefully controlling the location and extent of human, composerly intervention & decision-making.
  5. On a personal level, experiments in the arts seem to negate the possibility of revision, as such.  Any compositional action that might be seen as an effort to improve or refine the outcome is, for me, an independent experiment, an independent process.  For me in general, experiments need to be self-contained, closed on a local level, but open globally to larger trajectories of development or exploration that go well beyond the limits of any individual work.

The work as experiment / Experiment in Compositional Practice

The central driving force in my compositional method is an effort to create a process of discovery through composing.  Everything that I do as a composer, from the beginning of work on a piece to the end, exists to set up scenarios that allow me to invent material that I wouldn’t be able to imagine left to my own devices.  Principally these revolve around a many-layered set of restriction and constraint systems that, in a sense, control the available space for compositional decision-making.  When they work best, they force me into decisions – and, more importantly, into sounds – that I otherwise wouldn’t make.

My compositional method is driven by a ‘ground-up’ construction process – as opposed to ‘top-down’ – that forces all material (form, shape, scale, interaction between voices, gesture, dynamics, register, etc.) to emerge.  To take a reasonably simple example, structure in my work – tempo, sectional divisions, phrase lengths, meter – all emerge from the ground up from the intersection of local level processes, typically stacks of short loops of varying lengths.  These numerical cycles – which, not coincidentally, are constructed with random number generators – each spin at their own rates, and points at which the cycles click into nodes of alignment generate higher-level divisions and groupings.  (The more cyclic nodes that realign, the heavier the ‘weight’ of that higher-level division … and from there, those higher-level divisions form their own cycles, and their own nodes at points of alignment – this continues until the overall shape and scale of the piece appears.)  Critically, the form of a piece is not something I decide, as such – sectional divisions, durations, tempi all have a life of their own that I then have a responsibility to react to and ‘articulate’ through other lower-level compositional decisions.

There are other far more intricate examples.  My work with extended tablature notation, which extends back to 2004, has included extensive mapping of constantly shifting, sliding windows of available action on instruments, typically on multiple physical planes or with multiple aspects of an instrument’s mechanism.  I tend to work with large sets of independent and overlapping trajectories and vectors which control, or, rather, open up possibilities for types of action or gestural movement.  The intersection and collision of these various paths combine to create what are ostensibly phrases, though, as with the formal designs, the identities of these phrases emerge in quite unpredictable ways from the ground up.  As with the aspects of large-scale form, most of the details of these trajectories are determined initially through random number generators (almost always dealing with only 5 or 6 possible digits or values, at most).  And each layer of musical material – dynamics, registral shapes, gestural families, individual gesture types, fingerings, embouchure position or tension, bow position or pressure or speed, instrument behavior patterns, ensemble interaction tendencies, etc. – are all given independent values at the beginning and end of each phrase. Typically, I set up a very simple palette of possible gestural models (these are exceedingly simple and reductive in nature), with constantly shifting subsets of available options from within that initial collection.

In each case, the central thread is that underlying, and largely uncontrolled/uncontrollable forces put me as a composer in positions to react, to invent as a result of situations and restrictions that I hadn’t really designed, nor ones whose outcomes I would have been able to predict at the beginning of the compositional process.  It becomes clear why the unpredictable, vaguely Cage-ian use of random number generators becomes important:  the goal is to construct compositional systems that have clear shapes, clear identities and energies of their own, and to have those systems frequently, consistently put me in uncomfortable, unexpected, unplanned positions.  My compositional systems never prescribe what happens in any particular moment of a piece – instead they only ever control the number of options and the number and size and shape of the spaces within which I have the opportunity to invent.  The method is set up to force experimentation, to make me move away from learned gestures, shapes, figures, harmonies, forms, etc., and, as in the last example from the Second Quartet, the most exciting of these are the moments in which my options are quite severely limited or even self-contradictory.  It has been those most constricted moments in which the principles of experimentalism have become most immediately clear for me as a composer.

Experiment in performance

The obvious extension of the unpredictability embedded within the compositional process itself is a similar degree of unpredictability in sounding results.  My work since 2004 has foregrounded a substantial level of indeterminacy, from performance to performance and performer to performer.  These works take a highly prescriptive approach to notation, specifying physical actions and physical interactions with the instrument rather than resulting sounds, and as a consequence the relationship between score and sound is highly destabilized.

I’ve written and spoken fairly extensively about this indeterminacy in other papers and other venues, so in the interest of time I won’t repeat the details here, but I’ll instead focus on what, for me, is essentially ‘experimental’ in character about this approach. Most significantly, the near elimination of direct score-sound compliance in these works owes a substantial debt to the notational practice of American Experimental Tradition, particularly from Cage, but also from Feldman’s use of notation as a destabilizing force.  Because it is action that is specified in these works, and because this action is often intentionally vague/imprecise in its actual application in its interaction with the instrumental mechanism (most notably in the finger movements in the oboe/clarinet duo), and because there are frequently several levels of similarly imprecise activity overlaid on top of one another, the sounds resulting from these actions will, even for the player, often be surprising, unpredictable, and unstable.  (To take the wind duo Being itself a catastrophe, the diagram must not create a catastrophe as an example, the changes in embouchure tension or reed placement have a certain general set of tendencies, but the interaction with the instabilities of the finger movements will often create quite dramatically different results from iteration to iteration.)  This is, incidentally, also the reason why material in my work is so often transitional – there are almost never steady states in my work, but rather many layers of independent shifts between states, forcing the alignment and interaction of those layers to be even more unpredictable.

I often have the quite unsettling experience of being knee-deep in the process of making a piece without being entirely sure what it is yet.  I work extraordinarily carefully, and the working process is extremely labor-intensive, controlling mappings and densities and movements so that I really can track the questions that the piece poses, but I am very rarely quite sure what the piece’s real sonic and formal identity will be until it’s done, and even then that identity isn’t always clear.  The gestures themselves are designed quite intentionally to enable a process of discovery, to create unexpected and unpredictable results.  In the piece I’ve most recently completed, A painter of figures in rooms, we have found in early rehearsals and workshop sessions that each singer can follow the same set of notational instructions and generate quite dramatically different sounding results – indeed, one of the most exciting aspects of the piece is that the natural, personal attributes of each singer’s voice are magnified and intensified, undoing or at least unsettling the more learned, unified, pure, ‘choral’ sound that they are more accustomed to.  Equally significantly, the same singer might attempt the same gesture three times in a row and get three slightly different results, and it is here that exciting aspects of interpretation start to emerge.  As in, “Oooh, that one was cool, how can I get it to sound like that again?”

How do you know if you know what you’re doing?

This all raises some quite critical ethical issues for me, none of which I have definitive answers for yet.

First … What are the ethical implications when a composer puts something into a public sphere without a) knowing what it is, and b) knowing how, or even if, its experiments will work?  Because the experiment is the piece, and because its materials and shapes and identities only emerge through the process of making it, and the specifics of its soundworld only exist through its realization in performance, there is really no way to test or experiment before the fact (other than in a very generic, vague, tinkering, provisional manner).  That means, by definition, that the experiment is public.  We have an expectation, perhaps somehow one that lingers from a 19th century, (capital-R) Romantic notion that an artist has something to say, whereas for me the work I’m often most interested in is music in which the artist has something to ask, something they’re uncertain about.

This raises a further question:  what are the ethical implications of putting a performer in this situation?  How do I approach the compositional process and indeed even the rehearsal and performance interactions in a way that avoids making the performer something other than a guinea pig or lab rat?  How do I ensure that performers are collaborators in this process of experimentation, rather than the subjects of the experiment?

And one final general question:  To go back to the concerns from Ablinger from Michael Pisaro’s anecdote, ‘How do you know if you know what you’re doing?’  If the experiment is in the work, and its testing ground appears through performance, how many performances are required before a useful data set has been collected?  How many different performers?   How many different instruments or venues?  How much time is required before the findings of an experiment can be evaluated?  I know from my own experience that I have found certain works completely baffling and impenetrable – some of Ablinger’s work, for example! – that I now love deeply as a listener.  It took me quite a long effort before I came to love Peter’s Grisailles 1-100, and on first listen I couldn’t even manage to make it halfway through the piece.

What happens, then, when I have the same reaction to my own work?  What happens, sitting in the audience of a performance of one of my pieces, when I think, “I’m not really sure what to make of that.”  Our expectation is that it probably means that I don’t know what I’m doing, and perhaps that’s the case, but I don’t think it’s too much of a stretch to expect that composers of experimental music need the same acclimation time, the same space to process and react and understand a new work as the rest of the audience.  To take this one step further, and to offer something of a provisional answer to this question, I even feel that I have a responsibility to take this approach as a musician. To make something that I know will succeed, something whose results I can predict, seems to me to suggest that a) its materials are probably materials and crafts and methods and forms that already exist, and b) it is, for me, a strangely self-indulgent approach to composition (making the piece about the composer and a certain show of his or her skill, rather than about the field and expanding what might be possible, exploring new approaches to music-making and new opportunities for expressivity).

Though it perhaps sounds excessively bold, the approach that I’ve taken, its prioritization of experiment through process, its systems which force invention, which force me to wiggle my way out of seemingly impossible compositional problems, exist principally to ensure that I can’t rely on existing solutions, fulfilling what is admittedly an odd sense of personal duty to make my contribution to music-making one that is not primarily about my own skill or craft or identity but instead is about opening up new spaces for musical expression that indeed others might occupy better than I could.