Elements of Sonata Theory: Norms, Types, and Deformations in the Late-Eighteenth Century Sonata
James Hepokoski and Warren Darcy
Oxford University Press
It's long, and perhaps for guitarists whose chief interest is performing the book will seem like a tedious slog.
But this book is pretty amazing, and I'd say it's one of the best books about sonata forms I've read since the old stand-by of Charles Rosen. Hepokoski and Darcy bracket out sonata forms into five general categories and describe sonata forms as a series of variable scripts with flexibility built into them rather than as cast-mold Theme 1/Theme 2 forms. Particularly compelling is the category of Type 2 sonata forms they discuss, in which Theme 1 doesn't recapitulate but is subjected to substantial development in what they call the second rotation of the thematic modules. The archetypal Type 2 sonata would be the Chopin B flat minor piano sonata, first movement.
As a composer there's a lot in the book that I admit bored me because it's stuff I already think about on a sonata-by-sonata basis, but for students and performers all of the little details I consider redundant might be immensely helpful for being explicitly mentioned, such as all the ways composers could go "off script" by manipulating where and when (or even if) material from the exposition returns in recapitulatory zones.
This book also confirms for me a case I've been making elsewhere that there are more sonata forms in the early 19th century literature if you don't slavishly stick to just one definition of sonata form. I have some issues with the book but I won't get into those just yet. From an analytic standpoint my gut assessment is this is one of those books you have to read and be familiar with even if you disagree with it because it opens up some interesting interpretive possibilities for analysis of sonata forms.