The last two blog entries, which contrasted traditional Urlinie readings with those based on an Ursatz manqué, made an assumption that is typical of these modes of analysis but is questionable, especially for waltzes (or other small-scale pieces, such as songs). Far from being the staid forms that tie the chaos of the foreground to the eternal, these backgrounds are highly contingent -- fragile and easily disrupted. We've met some of the devices before: the stability of its tonal "reference point" -- whether upper voice or bass -- is quickly undermined if it is played as trio to another dance, or enhanced if it is given a trio and a subsequent da capo.
In the double-tonic numbers of D924 there is even another possibility -- that the strains themselves were detached from their companions and played in differing combinations. This is of course speculative -- an extension "backwards" to Schubert's improvisation practice -- but one does find such "floating strains" in Schumann's Papillons, for example, and the topically uniform character of most of D924 makes its both easier and more plausible as well.
I have assembled below a dance-sequence that uses the first strain of n3 as the refrain. More repetition or extension would probably be necessary, but even in this form up to three minutes of dancing would be available. The sequence is n3, a, b, a -- n2, all -- n3, a -- n5, all -- n3, a -- n4, all. The key sequence is c#, E, c# - E - c# - A - c# - A.
