somewhere along the line, schoenberg provocatively compared his music to mozart's, saying something about how it was more like mozart than not, or that everything he learned as a composer he learned from mozart, or something equally grand. being too literal myself, i took him at face value; once after an open dress rehearsal at severance hall i asked mitsuko uchida (who was playing the schoenberg concerto op.42) what she thought the similarities were; she clearly didn't care for the question but toed the party line herself and said there were a lot, without bothering to enumerate them - probably for the best.
whatever. in truth i haven't found it a useful comparison. in my experience of schoenberg's piano music, it's more useful to compare him to bach, in the way that the polyphony proliferates, and that all the individual voices are melodically or motivically designed. glenn gould's comparison was schoenberg to the franco-flemish masters a hundred years before bach, but bach was in the business of perfecting their arcane techniques anyways, so we'll go with mine.
flipping through schoenberg's 'theory of harmony' the other day - yes, my life is that glamorous - i found a fascinating but incomplete analysis of a chorale from bach's st matthew passion, no25 'Was mein Gott will, das g'scheh' allzeit' (What my God wants, should always happen). it comes in chapter 17, non-harmonic tones, and basically is about how what theory calls 'ornament' or 'passing tone' is sometimes essential to the harmony, or artistry, of the work in question.
schoenberg begins by commenting on how each of the four parts is written melodically, which as i said is pertinent to his own music (thinking especially of op.23). he points out that if you omit what we box in as 'passing tones,' the result would be something of very poor workmanship; he is emphasizing that they are determinative artistic choices. then he points out that the 'passing tones' in the inner voices of the first phrase form chords on their own, that are in dissonance with the tune but engage the poles of the tonality (I, IV, V), and he goes on to relate that to their rhythmic placement, etc.
if we took them as such in analysis, we would find an eleventh chord (e-g-b-d-f#-a) in the third beat, but nobody analyzes like that. i don't think. still, schoenberg's insight is a strong one because he showed how bach's harmony is much more nuanced, and in flux, then rules of theory can convey. i think we instinctively hear this, but to be aware of it is another thing entirely.
playing the preludes and fugues of the well-tempered clavier, i have often found that a so-called 'vertical' approach, that is, thinking in terms of chords rather than individual lines ('horizontal'), is counter-intuitively more useful in achieving the clarity of the individual lines, because after all they have to relate to each other harmonically in the end; they derive their inflection from that relation. his most pungent dissonances tend to be hidden and indirect, either with suspensions or space in range, but they contribute as much as anything to affect and atmosphere and mood.
in the grand coda to the prelude in f# minor book II, there's a tied e# in the middle voice that comes first as in the C# chord, the dominant, but when the other voices land on the downbeat, they form an augmented harmony that is fantastically dissonant (a-e#-c#) and seems to vibrate. yes, the e# easily resolves to f#, but to degrade that chord in our mind to a suspension or passing tone is criminal from an artistic point of view.
well, that's just another reason why i love schoenberg. not many people share that passion, and that's why i find myself blogging about his theory textbook at 4.30 am.