As the title suggests, Willner accepts the three-part Ursatz (enthusiastically, even—I am said to have "blazed an 'obbligato trail' with [the] three-part Ursatz, which allows for the structural descent of both soprano and alto" (2)); but he then expands the options to include the tenor and bass, though not in a consistently maintained voice leading grid (as if a kind of background chorale setting), so "not [an Ursatz] in which all four voices are equal, nor one in which the tenor part embodies a genuine structural voice" (2). The argument becomes a bit tortuous as he then asserts that "the background structure does indeed remain two-voiced at the very deepest level. The obbligato voices realized by the alto and by the tenor unfold a little closer to the surface than the fundamental two voices do. But in practice, as an aural and as an analytical experience, this is a distinction without much of a difference, at least in what concerns the alto" (2-3). I might complain that a distinction without a difference may not warrant a distinction at all.
Still, Willner grounds his adjustments in a compositional device that was especially important to 17th and 18th century practice: "Letting the Ursatz remain in a state of polyphony points to the dependence of all voice leading—from the foreground to the background—on invertible counterpoint" (13). He also makes a revealing comment about style features: "Despite the soprano’s prominence, it’s actually the alto’s descent that usually provides the scaffolding over which the thematicism of the piece rests, at least in the Baroque repertoire" (3). The analysis of the Allemande is guided by this idea throughout.
I will reproduce here only the background graphs from early in the essay, as these reflect the point just made above. The first graph shows a three-part Ursatz with a diversion by the alto into the tenor (arrow). I have added the red circles to bring out the alto, which—following from the comment above—Willner takes to be the primary upper voice.
The second graph demonstrates the source of this unequal pairing: the "obbligato soprano" would be a line of lower thirds under the primary voice, the "structural alto." (It is now also easy to see why invertible counterpoint is a factor.) My circled notes attempt to bring out these underlying thirds: F4-A4, E4-G4, F4-D"4", E4-C#"4", and D4 unison to close.
Postscript: "Examples [of minor-key ascending Urlinien, despite the odds against them] can be found throughout Bach’s suites, sonatas, and partitas for various instruments. Most movements of the D minor Suite for Violoncello Solo, for instance, close with a motivically charged ascending Urlinie" (14n20). Here I will respond with the same quibbling distinction Willner made about the three-part Ursatz and the background—I'm not at all sure those ascending cadence gestures belong to the largest-scale melodic features in the Bach cello suites, despite their position in the final cadence. I have written so far only about two major-key cello preludes, Eb and G: link; link.
