I’ve been writing little analytical posts or random theoretical musings on Twitter on and off and have been thinking it might be a good use of this space to put some of these thoughts here.
I was just working on the Bach chorale BWV 38.6 with a student on the weekend:
It’s probably the most dramatic example I’ve seen of a chorale where Bach has tried to graft a functional harmonization onto a modal hymn melody. The melody is very obviously in E Phrygian but Bach has harmonized it in A minor despite the fact that only the third phrase of the melody lends itself at all to a tonal centre on A. The hymn melody begins on a B and ends on an E. Gs are always natural in the melody but often raised to G# in the harmony parts to make the harmonization function in A minor. The first, second, and final (!) phrases all end with half cadences on E (V in A minor). The third is the only one that ends with an authentic cadence. The fourth modulates to G, which is obviously the relative major of E minor but is an unusual key change for a piece that is otherwise in A minor. I’m not sure it even works completely but it is interesting that we get the only authentic cadence in the home key on (trans) “He alone is the good shepherd”; we also get an authentic cadence in G on “who can free Israel” but are denied resolution on “although there is much sin among us” and “from all his sins”.