The evolution of P/F vs T/K raises important questions

[Mauro]

I guess we'll have to agree we disagree on these points, and live with that!

OK, but what is your theory about p/f?
All the best, --stolfi

I'd be happy to have one! The only things I can say is p/f/k/t are probably in the same glyph class (which may also have cph/cfh/ckh/cth as members and, less probably, also ch/sh), and that p/f and k/t form two distinct sub-classes. They (and expecially p/f) also have a peculiar spatial distribution on the pages, and their proportions change along the manuscript. Unfortunately, I can't go beyond that.