I was contemplating the opening progression from Bach's harmonization of "Christus, der uns selig macht" a few weeks ago (#307 in the Riemenschneider). What struck me at the beginning was that when I first went through it, I heard the first measure as a I - iv (a rare but not unknown progression for the Baroque) in spite of the key signature. Naturally, by the end of the 2nd measure, my sense of gravity had shifted, and what I thought was a I turned out to be a V. Naturally, I opted for thinking about the passage in terms of a Charles Smith multivalency approach. So, I went back thinking about the first chord as simultaneously I and V. Curiously, I couldn't hear it as I-chord anymore. Even with great effort to hear it as a I-chord, I could only achieve a I/V hybrid. So:
Is multivalent meaning conditioned by context in such a way that something can be honestly multivalent (as opposed to being multivalent through misunderstanding) in one hearing and then lose its multivalency on subsequent hearings?
Also, what is the story with the chorales ending on half cadences? I will give the caveat that I have only seriously looked at 311 out of the 371, but of the chorales in a-minor that start on some sort of E7 chord, #3, 10, 21, 81, 190, 198, and 307 all end on half cadences. The exception being #57. Does the V - in spite of (or maybe because of!) its instability have some sort of gravitational pull that supersedes the tonic? I think, given the number of examples that I have listed, we can safely exclude the issue of "text expression" as a significant motivating factor.
(Incidentally, #307 is a very clever chorale. At least to my ear, every cadence (including the last one!) in the chorale sounds like a half-cadence with one exception. The exception is not in the tonic key. It is most definitely the result of some "vagrant" chords interfering.