Mike - first off, super job on this excellent piece. It's really interesting reading. And your final point, that very few other bands including the Grateful Dead routinely change keys during a jam is one that I think we all intuitively know, but it's nice to hear that sense confirmed by an expert.

I'd be curious to get your take on earlier periods of Phish jamming, specifically '92 - '95. While the band did modulate jams at this point, it seems to me that a bigger aspect of the jamming back then compared to '97 thru to now was rhythmic changes. For example, in a Bowie, Antelope, or SOAM back then, you have frequent brief rhythmic change ups that seem less common today. At that time, Trey was clearly still driving the jams, and Fish seemed keenly dialed in to Trey's sudden and abrupt suspensions or other temporary changes to the underlying rhythm. I sort of think of this period as a time when the band (or at least Trey) was still considerably under the influence of Frank Zappa's compositional style. With the more cohesive, band-wide jamming that emerged in 1997 and on, it seems that sudden rhythmic changes are more difficult to pull off without a clear leader. On the other hand, the modulation of key, as you clearly note, can be signaled by one band member, and adopted by the others once Mike buys in to the change. Anyways thanks again for the great and detailed lesson in Phish music theory.