Measures 33-51: I64, V7, iii, I64, I6, vi, ii, i, IV6, IV, ii, V, iii, vi, IV, vii065, vi, vi6, IV64, iv65, I64, I, IV, iv, N, N6, V, V6, iii, I, V, I, I, IV64, ii042, I.
The climax of "Die Mainacht" is in measure 45 where Brahms emphasizes the word "heiser" by placing a Neopolitan harmony underneath. This makes it sound more "burning" and upsetting. Additionally, Brahms intensifies the shock quality by creating a descending vocal line for the climax- an unusual move. I think most people would disagree with me and say that measures 39-43 are the climax, but I think that the lighter harmonic color doesn't hold a candle to the climax coming in measure 45 with the angry neopolitan. Earlier in the piece, you hear many d naturals- Brahms way of preparing the listener for the nasty ending with the neopolitan.
The first root position tonic triad on a downbeat in this section is in measure 38 with the dominant harmony. I think Brahms does this to allow the poet to state his question, although I don't think he wanted the singer to lose the phrase he's been creating. As I mentioned in my earlier blog, I don't really believe that the phrase ends until the piano is done in measures 49-51. This music is all about the connection between the pianist and the vocalist, so phrases can be intertwined and rely on eachother. (That's also how Brahms can get away with making phrases so long, you wouldn't see phrases like this in Handel.)
I disagree with the workbook saying that "it is absolutely necessary for performers to determine from the accompaniment where true phrases occur." I agree if we're only talking about this period, but like I said, I don't think Handel would've cared that much about the accompaniment of an aria. In the case of "morgenrot" - it is not the end of a phrase, Brahms communicates in a number of ways: 1. The moving piano part implying to move forward 2. The word "morgenrot" means rosy morning- the morning implies a new day, if it were night, maybe it'd be the end of a phrase, the poet's obviously got more to say about his morning. 3. The harmony is not cadential or even close to sounding complete- Brahms has a reason for doing this.
I would not interpret measure 44 as a phrase ending. Once again Brahms creates a moving piano part with moving harmonies, and it wouldn't make sense for someone to state a complete thought that way lyrically. I think Brahms just wanted this to be a forethought to measure 45. As if the poet is at first stating his revelation quietly to himself in measure 44 and then crying for attention in measure 45. I would move forward in measure 44 and take more time in measure 45, as a performer.
The piano postlude is VERY important, structured mostly of diatonic harmonies, except for the diminished ii42 right before the final I. This piano part finishes the "wandering" where the singer left off. The movement of wandering paints the scene for the listener while they're still thinking about what the poet said. The listener is then given extra time and reminded of the bitter ending by the mixture chord with the fermata, right before the cadential return to tonic. If I were the pianist, I'd really "ham it up" in this section, especially since you don't have to worry about the singer anymore. I would use a lot of rubato and really emphasize the fermatas in the last measure.