The G-flat major prelude is the 13th in Opus 11, a set modeled closely on the Chopin Preludes, Op. 28. Although Swinkin does considerably more with this, I call attention only to the opening segment of analysis, which finds an ascending Urlinie. First of all the score, which reflects, along with the composer's strong absorption of the Chopin model, an unusually strong left hand (acquired as the result of a time of practice limited by an accident to his right hand).
Next, my harmonic reduction, whose main task was to dispose of the many accented neighbors:
Now essentials in Swinkin's analysis example. Note that he has produced a harmonic reduction that is more diatonicized than mine (mine in other words is an intermediate stage between score and Swinkin's graph):
And mapping the notes of his analysis back onto the score, with an assumed tonic root-position chord and ^5 above before the composition begins. What this claims is that the design only arrives at clarity through the cadence, and thus allows (or requires) a retrospective reading of the beginning. By the later 19th century, such processes were commonplace (vide Debussy).
Finally, for reference, a less adventurous analysis that takes the prominence of ^3 as a clue to the background. This is an entirely plausible reading that misses almost everything of expressive interest in this prelude.
Reference: Jeffrey Swinkin, "Performative Analysis: Theorizing the Interpretation of Tonal Music." PhD dissertation, The University of Michigan, 2013.
