Sometimes this opportunity is realized through some grander compositional or theoretical project, figuring out, for example, what pitches might do for me if very flexible and accurate intonation were reliably available at little cost, OR calculating all the possible subdivisions of metrical feet in some large number of possible measured metres, and then figuring out how these weighed against one another in possible phrases, OR building a practice of harmony based on smooth voice leadings from an octave divided into equal parts OR trying to built a better orchestra... But, as often, the opportunity involves taking those idle moments that emerge with deep doubt and simply doing musical exercises, a return to the most zen-ish disciplines of apprentice days : for me, these are usually contrapuntal — species exercises against a cantus firmus, or canon writing (at unison and intervals, perhaps inverted), or bits of fugal answer. (Others probably find more use in harmonic exercises, but — for better or worse — I've always experienced harmony more as a consequence of contrapuntal lines than of sequences of chords.) It's important for me to do as much of this in my head as directly on paper, and "in my head" usually means while walking (hence one valuable quality in having Terrier Mutt Lucky, the Composer's Best Friend around, to force me to regularly get up and go walk about), but however it gets dictated, I do have to get back to a keyboard and play it through as soon as possible, or the exercise is simply not complete. These periods of intense exercise usually include some intense score reading; at the moment, it's one Contrapunctus or Canon from The Art of the Fugue each day at the keyboard, once in the morning and once at night, cycling through the collection, no matter how little my fingers cooperate these days. I think the choice of music to play is actually not so important — last time around, I did some serious gamelan playing, next time, who knows, maybe I'll take up the viola again — the important thing is simply to have such a regular, disciplined practice in a repertoire with such clear constraints that the musical imagination is given a regular opportunity to wander.
The other opportunity this moment offers is for reassessing music history a bit. (Sometimes I think my concerns with music history and ethnography are a bit perverse; they're certainly not widely shared with my colleagues, but please grant me this one perversion!) I've been incredibly disappointed in the most popular or prominent versions of recent music history that are on offer (although he only really covers the first half of the 20th Century, William Austin's survey is so much better than that of either Ross or the last two volumes of Taruskin's Oxford), but doesn't that simply make it more interesting, or even urgent, to formulate your own version? On the one hand, I've become somewhat fixated on the notions that continuous memory of musical practice really doesn't extend before the last quarter of the 18th century and that the introduction of sound recording in the early 20th actually handicapped memory rather than reinforcing it, that the loss of the play of tropes and figures and affect that predates the "classical" era was a real one and worth serious investigation, and that many of the breaks our textbooks and semester schedules insert into repertoire, the turn of the 20th century in particular, were not only artificial, but actually deceptive. I'm only at the start of these thoughts and I don't know how any of these ideas will play out compositionally, but they surely will.