I like the musical metaphor.
Behavior IS in some basic ways music. We have notes (acts), melodies, chords, traspositions, and so on.
Human speech is music. That is something that formal linguistics can ignore, to its peril.
Here is one obvious thought on context. In human speech there is the well known phenomenon of "co-articulation", where the very pronounciation of a sound varies as a function of other sounds the precede, and follow. Niftily, the hearer of co-articulated speech can re-process these changed phonemic components, and bring them back to their original template. Context determines components, in a sense at least.
Context, in a causal sense, is probably more important that can be revealed by reductionistic models. "X" has one effect in context 1, and a different effect in context 2. X and Y may even "take over" and replace each others effects, within certain limits.
[Here is an unintended one on context....my computer just got cut off line.......took my thoughts with it....so I'll stop short. Music is a lovely analogy for living processes. The key question concerns homology at a deeper level of organization. Would be nice to pursue that....if and when my computer comes back!]