end condition

December 30, 2016 - news / workbook

Since beginning to tag my pieces, I’ve been looking at which of them might qualify as games. Perhaps unsurprisingly, it’s very few: only three. Although I’ve been investigating games a lot over the past year, I’ve not really translated some of the core requirements for the pieces to fit any of the common definitions. I find Chris Crawford’s taxonomy a useful one. Most of the pieces I’ve made are interactive, and they often feature competitors who make ‘attacks’ on other players – or at least influence their actions. There are also activities and rules that constrain behaviours, but very few of my pieces include explicit goals, other than following the instructions and trying to make interesting and engaging choices. There is no purpose that motivates players intrinsically. In the three pieces that do have a specified goal, the sense of completion on achieving it is a lot clearer. With all the other pieces that use interpersonal interaction and cueing, the sense is closer to open-ended play. Goals provide completion, purpose and, perhaps most usefully, a meaningful end condition. Without this, endings are essentially arbitrary (stop after 10 minutes, stop when it feels like the end, stop when you hear me make this sound etc.). My hope is that goals provide a sense of direction, intent, and resolution that is perceivable by those watching the performance. So there is an empathic experience, whereby listeners have an understanding of what the players are doing.

The first of these pieces I’ve made that I feel begin to become gamelike, all voices are heard, presents players with a pool of sounds which they sequence independently. Each repetition of the initial sequences is adapted by the players individually with the aim of achieving a unison result. Each instance of the process completes when the players, and hopefully audience, are satisfied that everyone is in agreement. There is a purpose that leads to an identifiable configuration of actions.

In in which one thing depends on another, players must work cooperatively to associate a series of 20 words with 20 sounding objects. They begin with no associations established and then gradually form object-word pairs by teaching, reinforcing, and testing each other. The piece completes when both players are secure in the associations and can successfully test each other. Again, there is an observable learning process that is tested by an end condition, and we see the players set up the associations and make (many) mistakes along the way.

In we gradually have more things to do and fewer things to say – part of the things to do series – players have an increasing set of sounds to make and a corresponding decreasing list of cues to give. When one player runs out of cues, they hold their hand up and the piece finishes.

There are two different end conditions here: consensus and a race. The first two pieces are co-operative games in which players work together to achieve the end condition (unison playing; a stable set of elements). The third piece is a simple race (with attacks), with players adopting strategies to maximise their chance of being the first to finish. These conditions are effective and clear, and provide intrinsic motivation (working co-operatively; beating opponents) providing players choose to engage (a necessary assumption for games to function). There are many other game end conditions to explore of course, such as running out of time, collecting the most of a particular resource, or being the final player to be eliminated. My new orchestra piece is likely to work around player elimination (it’s kind of like Haydn, but with live decision making) for instance, with players stopping if they make mistakes in the responses they are asked to make. I’m still working this out at the moment, but it’s looking like a likely goal.

Over the next couple of months I’m working on motivation and purpose in game compositions, and indeterminate music more generally, both in my own pieces and as preparation for some papers I’m developing. I’m curious about the general lack of ‘win states’ in indeterminate music and the fact that as composers we don’t tend to express goals that embody a purpose. They might state a situation that results in the piece ending, but often this is arbitrary so as to fulfil this function in a practical way. The process behind the piece, the space in which the players play, is not necessarily connected to this explicitly. It seems like a missed opportunity to create tangible and communicative frameworks for music.