This is another prelude and fugue from the cycle for which we have a video performance. In this case the prelude is melodically simple but harmonically complex. It makes sense to provide a schematic analysis that doesn't attempt to get too involved in the harmonic activity of this prelude, even though a theoretical analysis could be a lot of fun to do. This is a series of blog posts, after all, and for blog posts it should suffice to give guitarists and theorists a foundation for more detailed score analysis at their leisure.
This block chord chorale is 39 measures long and is predominantly in four measure phrases in common time. Its form is probably best described as a kind of rounded binary form that can be mapped out as follows:
A (0:06 to 0:35) measures 1-8
A' (0:35 to 1:00) measures 9-16
B (1:01 to 1:24) measures 17-24
A'' (1:25 to 2:23) measures 25-39)
Although the phrases are in four-measure increments the larger phrase units tend to be eight-measure phrases. A is firmly in B minor while A' shifts into D major with the promise of a resolution back into B minor which is deceptively resolved not to the submediant key as would be typical, but to an F major added sixth chord that is the dominant to B flat major that is the tonality that the B material moves within. On the whole all the melodic activity is reliably diatonic until we get to this B section, even if the harmonic support is often chromatically divergent from the "home" key. The B material closes with a presentation of the G major chord that would be the "expected" submediant chord of a conventional deceptive cadence--we get the expected chord for such a deceptive resolution but it's postponed until the end of the B material in this rounded binary form. A short chromatic rising passage leads us to a recapitulation of the A material as A''. The initial idea returns but the phrases are shortened and more halting, the rhythmic values more expansive. We're obviously winding things down. It's worth noting, for those into this kind of analysis, that the return of the A material at measure 25 corresponds closely to the Golden Mean of the prelude.
As to the mood of this prelude, which I like, it gives me the impression of a kind of Scriabinesque take on the old American song "My Funny Valentine".
The fugue is introduced in the film by a change from a black background to a white background in this film, which makes it very easy to point out when the fugue begins.
This is a serpentine 6/8 subject that's four measures long and is dominated by a chromatic descent that winds up into eighth notes that lead to a countersubject in the uppermost voice as the alto enters at measure 5 with a tonal answer that affirms the dominant (2:31). The countersubject appears in a transposed form at measure 48 (2:36). It's worth noting here that tonal answers modify the subject in some way to adhere to the prevailing tonality of the exposition. When a subject starts on the dominant scale degree and you don't want to change the character of the subject's linear activity then you may want to provide a real or tonal answer at the subdominant, but if you want an answer at the dominant to affirm the key of your fugue exposition and you are willing to "bend" the contour of the subject then tonal answers are more satisfying. It's possible to favor tonal answers (Bach) or real answers (Handel) depending on preference. Neither is necessarily easier than the other to master. In this case Koshkin opts for a tonal answer at the dominant in which the subject starts on the third scale degree. As noted previously, he has a countersubject that is transposed for the appearance of the third and final voice in the texture which arrives in the tenor/bass voice on F sharp just as would be expected.
We move from the end of the exposition into development at measure 52 (2:42) and right away Koshkin shifts to F natural, signaling that the chromaticism of the subject opens up possibilities for abrupt modal mutation and chromatic shifts into remote keys. Despite the three voices of the exposition by its completion this fugue will be characterized by two voiced textures. This is not a weakness in a fugue such as this, given the activity of the subject and its brevity. As George Oldroyd put it in The Technique and Spirit of Fugue, if memory serves (and it may not) one of the mistakes music students make in writing fugues is assuming that once all three or four voices in a fugal exposition are introduced every last one of them must be active constantly throughout the duration of the fugue. Anyone who has surveyed the fugues of Bach or Buxtehude or ... just pick someone, you'll find they rarely ever did that. So you don’t' have to, either. Koshkin demonstrates a firm grasp of contrapuntal writing by recognizing he doesn't need all three of his voices active all the way through this fugue.
The first middle entry appears at 2:50 and is in the very remote key of C minor at measure 58. After another episode the second middle entry appears at measure 70 in the key of F minor, yet another wildly remote key from the initial B minor key we started in. A very brief transitional episode takes place and by measure 75 (3:13) we're already in a new middle entry in E minor. Koshkin clearly grasped that one of the things you can do with a fugal subject like this one is move quickly from middle entry to middle entry with little episode transition thanks to the chromaticism inherent in the subject. This passage in the fugue is also noteworthy for the way in which the F minor middle entry places the subject in the tenor/bass voice while the E minor middle entry places it in the alto voice. He also features the countersubject in the earlier middle entry and omits it in the later middle entry, which is something else you can do when writing fugues.
By measure 3:20 we have still another middle entry at measure 80, in G minor. This is followed up by an extended modulating episode that downshifts with descending sequential activity in linear terms but which culminates in an "uplift" at measure 92 (3:35) in a new middle entry in A major.
By measure 100 (3:48) we have a new middle entry in which the subject appears to have returned in B minor but it's not B minor. Though the F sharp appears at the top of the treble clef just like it did at the start of the fugue it is floating above a harmonic tenth of E at the bottom and G in the upper part of a dyad. The fugue subject has returned but as the ninth rather than the fifth, part of a middle entry that is more readily identified as E minor than B minor. That's something else that makes fugue subjects with a pronounced chromatic character so much fun, you don't have to bring them in at the tonic or dominant if you know you can transpose them in some other way. If you want an example of a fugue that runs with the possibilities of a chromatically ambiguous subject as being able to appear at any factoral relationship in a chord (tonic, third, fifth, seventh) go over and listen to fugue no. 2 from Rodion Shchedrin's cycle of preludes and fugues. Go over here
His fugue in A minor starts at 2:55.
Now, back to Koshkin's fugue.
The E minor middle entry with the subject at the ninth leads to an extended episode that can be thought of as a preparation for a "real" E minor middle entry that finally arrives at measure 112 (4:02) with the subject appearing at the fifth of the chord as we would expect. There's even a quasi-stretto going on in which the subject in the treble strings gets a semi-canonic response by a transposed form of the subject in the middle voice. This overall rhetorical flourish is, by the way, almost a Haydnesque trick, giving the subject in a middle entry at an unexpected interval in a harmony and then preparing for a payoff in which the subject appears at the "correct" register later in the work. This is followed by a flurry of sixteenth note scalar runs that begin to get support from an F sharp in the lower register. As wind ups to fugal climaxes go this is a class decorative series of runs on a dominant pedal point; this is exactly what makes sense for a guitarist to do in a fugue in B minor and it makes for a satisfying climax in which the fugue subject arrives in its original B minor as B minor at measure 123 at 4:18.
As the subject reaches its end it gets interrupted by florid runs that preempt a full resolution of the subject in harmonic and linear terms. Koshkin goes through a series of evaded cadences to build suspense. He finally gets back to the motto of his subject at 4:42 (measure 137) where the subject becomes fragmented and the basis for winding down coda that reaches an E major chord (though it has an added sixth and no fifth) at measure 147 that turns into a florid scalar ascent culminating in a natural harmonic that is followed by a subdued piano tonic chord close.