Study 22 from Fernando Sor's Op. 29 set of studies for guitar (study 18 in the Segovia collection) is unusual in that it is in the key of E-flat major. There are very few pieces for guitar that are written in this key for obvious reasons. As with Sor's study 13 in B-flat major from this same set we have the similar challenge of playing smoothly with fewer open strings at our disposal. I have always avoided this study until now for those reasons. Musically it is top notch as with most everything Sor composes. Let's take a look at some of the interesting features.
The edition I used for this analysis is a 1924 edition published by N. Simrock, Berlin.
As you can see from the analysis this study contains some rather dense and complex harmony usually associated with some sort of chromatic voice-leading. The two most complex harmonic passages occur in measures 29-30 and in measures 80-81. If you play only the outer two voices in measures 29-30 you will hear the interval pattern that generates the harmony. Usually it involves a pattern of a tritone resolving to a third (tenth) in contrary motion. The one exception is the perfect fifth (D-A) in the outer voices on the last eighth of measure twenty-nine (D7) resolving deceptively up one half-step to Eb major before the pattern resumes again, ultimately ending where it began on Bb major in measure thirty-one. Notice the use of secondary leading-tone diminished seventh chords as well as secondary dominants.
The second sequence (measures 80-81) is even more complex. Sor starts with three descending tenths, the middle one (Gb-Bb) being chromatically altered to form a minor tonic (i). This Gb is not found in the Segovia version, at least not in my edition. This is followed by the tritone to third (tenth) sequence with the final tritone at the end of measure eighty-one resolving outward in contrary motion to a sixth to ultimately bring us to the F minor (ii) chord. This F minor in first inversion prepares the dominant (Bb) leading to the cadence in Eb major (tonic) in measure eighty-three. Again notice the extensive use of secondary leading-tone diminished seventh chords as well as secondary dominants.
A couple of other things to note harmonically are the common-tone diminished seventh chords that are heard in measures twenty-one, fifty-five and seventy-two and the augmented triads in measures twenty-one, fifty-seven and seventy-two. These chords are always best understood as a product of chromatic voice-leading. You can easily see the chromatic voice-leading in these passages as Sor is always aware of the way in which the voices move according to the conventions of counterpoint handed down by J.J. Fux and J.S. Bach. I think you will agree that measures fifty-two through fifty-nine are really a prolongation of the dominant (Bb) as indicated, with the chords in parentheses functioning as passing tones.
The form is ternary A-B-A (coda) as is typical of many of Sor's longer studies. The A section (mm. 1-40) is defined harmonically with the move from tonic (Eb) to a cadence on the dominant (Bb).
The B section (mm. 41-59) is somewhat developmental in that we have more keys represented including the supertonic key of F minor. You can see from the analysis that harmonically we have essentially a big ii-V progression with the section ending with the dominant prolongation discussed earlier before returning to the A section with the resolution to tonic in measure sixty.
The final A section (measure sixty to the end) begins with a restatement of the opening six bars with the seventh and eighth bars modified to avoid the tonic cadence in order to begin a coda section on the dominant in measure sixty-eight concluding with the final cadence on the tonic.
There were some discrepancies between the Segovia edition and the edition I used for this analysis. First as mentioned earlier the Gb in measure eighty of my edition is a G-natural in Segovia’s. Secondly, I also noticed the E-natural on the final eighth note of measure eighty-one (soprano) is an Eb in the Segovia edition and seems to be fingered as such changing the harmony from Edim7 to Eb7 which could still resolve to F minor but less emphatically. Thirdly, the Ab’s in the middle voice in measures forty-two and forty-four are changed to A natural in the Segovia version changing the harmony from F minor to F major which is clearly wrong as F minor is certainly the implied key with the inclusion of the Db at the end of measure forty-two. Curiously the A-natural is followed with an Ab in the soprano in measure forty-four, beat two of the Segovia edition. Evidently all these discrepancies have been discovered and corrected by others including David Tanenbaum whose recording of this study is true to what I believe Sor intended.
All in all a very interesting piece well worth the effort. I hope you agree. The pdf file below contains the analysis. Thanks for reading and enjoy the piece.