Analytical discussions of enharmonicism often engage enharmonic transformations in the sense of modulatory processes involving changing the meaning of one or more tones upon shifting to another key. However, one may also point to constellations where two enharmonic interpretations coexist simultaneously, leaving the ambiguity essentially undecidable. In this paper I engage two striking examples of undecidable enharmonic ambiguity in late tonal repertoire, both involving one identical, perceptually multivalent sonority.
In the second Intermezzo from Op. 117, Brahms introduces the sonority in question at m. 8, ostensibly serving as a chromatically altered deceptive cadence, but notated diatonically as an incomplete seventh degree with an added non-chord tone. Drawing on Arnold Schoenberg’s theory of harmonic progressions, the two readings engender diametrically opposed interpretations—one involving “descending” root progressions, the other “ascending” ones—in a way evocative of Escher’s famous lithograph.
Schoenberg uses the same sonority as penultimate chord of his tone poem Pelleas und Melisande, notating both pitches—C-sharp and D-flat—simultaneously at different registers. This puzzling double notation may be traced back to a thematic liquidation process—the C-sharp is derived from the “fate motive,” the D-flat possibly from Melisande’s theme.
While analytical approaches applying reduction of the surface level tend to view ambiguous constellations as local deferrals of an eventual, unequivocal understanding of the music, I argue that undecidable ambiguity of the type discussed here is analytically irreducible. Intriguingly, Schenker’s unpublished graph of Brahms’s Intermezzo Op. 117, No. 2 arguably reveals his difficulty in addressing issues of structural ambiguity.