Dynamics

I've always had a problem with dynamics.
How can I possibly have a problem with dynamics?
What am I talking about?
All of my scores have dynamics until about 2001 and then... some sort of crisis?
Sorry, that sounds ridiculous. You can't have a crisis about dynamics.
OK, so it wasn't a crisis, merely a realisation that I didn't know what I was doing.
I was plastering these markings around but they weren't actually contributing anything to the piece other than making the score 'look right'. There was nothing as such wrong with them. They just weren't really part of the score in the same way that pitch, rhythm, instrumentation, etc. were.
I was quite content to just start putting one dynamic marking at the beginning of an entry (and sometimes at the beginning of a piece) and just leaving it at that. If there was phrasing, the performer can work the dynamics out for themselves.
So that worked to a limited extent. I started my PhD and was producing a large piece for the London Sinfonietta (Lovesongs) and I couldn't even commit to dynamics at the beginning of each entry. Did I really want a monochrome colour throughout? No, but there was phrasing which the performers interpreted.
I felt paralysed. I could go through the score and 'improvise' dynamics but how could I convince anyone that I really wanted that crescendo if I felt that, on another day if I had eaten a banana for breakfast instead of toast, it could have been a diminuendo? So the Sinfonietta received a score without dynamic markings (well, there were a few).
Everything else in the score was meticulously notated. (Carl Raven put his finger on the problem about another piece written at the same time when he said that there was a mismatch between the almost aleatoricism of my dynamic markings, and the precision of every single other aspect of my notation.)
F grilled me consistently over the course of the first three years of my PhD about my relationship with dynamic markings but I couldn't get out of this hole.
So I went back to an earlier piece (disiecta membra) and came up with a series of structurally determined processes to generate dynamics and articulation markings. The whole piece was generally shapeless in the form I'd left it in 2003, so I set about creating a process with dynamics and articulation that meant that page thirty was different to page four (which wasn't really the case before I enacted these processes on it! There is a high degree of self-similarity in all of the pitch/rhythmic material that didn't quite work).
Working on this piece in this way unlocked something. I suddenly felt able to attack Lovesongs and create meaningful dynamic markings that really brought each individual line and gesture alive. Now, looking at the score, it is finished. Without them, it was flapping in the wind, a perfectly serviceable piece of music yet somehow unsatisfactory. F called it a torso.
Last night, talking to him, I think I figured out what changed. In creating almost completely automated processes to generate dynamics, I discovered a means by which dynamic markings could transform material into something more significant. At almost every level of my compositional practice, I have to be transforming things, transcending their existence as notational cyphers into music. And suddenly dynamics could also fulfill this function.
F sounded a note of caution.
He said that some of my ideas about flexible dynamics were potentially interesting in the right piece. So that's something to return to. I even have the basic idea for the piece...