Here is Mozart's K545. The melody goes from A to G at the change of harmony twice and the accompaniment does the same. Is this bad voice leading? Or rather: Why is this NOT faulty voice leading? Is it that voice leading between the "inner voices" (not bass) of the accompaniment and melody are voice led seperately? Or is it that the accompaniment changes from two to 3 voices that compensates for this?
The parallel motion between the melody and an inner voice in Mozart's K545 is not faulty voice leading for several reasons.
In essence, while strict parallel motion rules apply to certain contexts (e.g., chorale writing), homophonic textures in classical piano works have more flexibility.
This is not strictly polyphonic music. This piece switches between voice over Alberti bass, Voice against Voice, Voice against two note harmony and Voice against three note harmony quite a bit. And Voice leading is a bit of a moot point when you do not have clearly defined voices. And while this can be seen as a hidden parallel motion in some sense it does not really matter, as that fifth on the secondary dominant is only there to beef up the chord. It it not really perceived as a distinct voice.
Still of course caring about voice leading still does make sense here, but not to the extent where this would be a fault.
This is (as you say) melody + accompaniment, not a choral texture with 4 equal voices. Voice leading is not really much of an issue in this style. Maybe between melody and bass line. But don't look too hard for 'voices' in block chords.
See my answer in this thread:
Earlier in. K 545, I think starting at m. 5, there is a passage of simple scales in the RH. A pianist once told me those scale steps are not the melody. The point of the scales is really figuration to create rhythm. The "melody" in passages like that will come from a harmonic reduction of the figuration texture. In that particular passage starting at m. 5 that reduction is just a descending scale, in thirds, in whole notes.
Compare that figuration texture with passages like the opening four measures or the melody opening the exposition section in G major at m. 14. Those two passages are "proper" melodies. You can harmonically reduce them, but you would also pay close attention to their note by note motions in regard to voice leading/relative motion.
Getting back to the passage you posted (mm.26-28.) You first need to decide whether to treat it as a figuration texture or a proper melody. There is no right or wrong way to make that determination. But you can use some kind of common sense test. Can you sing or whistle it? Probably not, so it's best to consider it figuration.
Harmonically reduce it before considering voice leading. You can do this several ways:
Some combination of those three usually come into play making a reduction.
Also, use your ears to make broad assessments of harmony of phrases. You should hear that the example you posted is clearly cadential harmony. It's a rhythmic figuration, using brilliant fast arpeggios in sixteenth notes, of simple tonic/dominant cadential harmony, just before the final measure of the section.
The reduction is something like this:
| G G G F# | G G G #F | G | G - - D | G - - D | G
If you want four part harmony, to better explain the LH chords, rather that the two part bass and treble reduction, the change over the bar line is:
D7, FA descends to tonic's third MIA and bass D move to double the tonic chord root GThe melody goes from A to G at the change of harmony twice and the accompaniment does the same.
Depends how you reduce it. The historical topic to look into is diminution. Basically that is splitting long notes into smaller ornamental notes. The final two sixteenths of the bar, F# A, can be viewed as a diminution splitting of a hypothetical F# eighth note. Regarding the F# as the main harmonic tone makes a lot of sense given it is the "missing" tone from D7 with D A C in the LH.
Is it that voice leading between the "inner voices" (not bass) of the accompaniment and melody are voice led seperately?
Distinguish ornamental notes from harmonic notes. The ornamental figuration is rhythmic. It provides a burst of sixteenth note activity. That rhythm could have been realized in repeating notes, but that would be harder to play and not matching the figuration of the rest of the music. The sixteenth notes logically pass through the tones of the chord, because this is homophonic (not polyphonic) music, but that isn't a reason to elevate the importance of those fleeting pitches beyond F# G in the RH to essential harmonic tones.
Or is it that the accompaniment changes from two to 3 voices that compensates for this?
Some of this can be regarded similarly to instrumental doubling where the LH and RH can be thought of as two instruments. This will account for many, many supposed pitch "doublings" and "voices" that seem to get added or drop out of the texture.
One final thought. This harmonic reduction business can be confusing some times. The RH part, sixteenth notes and all is the melodic line. You would analyze it melodically for contour etc. But that is melody. When you ask about voice leading and harmony analysis we need to shift away from melodic details to see the harmony. It doesn't mean there isn't a melody in those sixteenth notes. We just don't look at all that melodic activity to assess the harmonic picture.
Succinctly, the last sixteenth note A in the right hand is just a filler, due to Mozart's quintessentially Classical-period piano sonata style. So I agree with @MichaelCurtis's harmonic reduction argument. The note that matters for the perceived cadence should be the preceding sixteenth note F#, thus avoiding the doubling of A->G resolution in the LH.
We then have:
F# -> G C -> B A -> G D -> G
which should be fine since it avoids parallel voice leading in the V7 -> I cadence. Play that cadence in the piano. Notice how the top 2 voices feel very satisfying (top leading tone goes up to tonic, and the C, the most important note in the dominant 7th chord D7, goes down to B, the most important note of the G chord).
That cadence is following a "textbook" method. See this theory textbook section 27.3 Voice Leading the V7 to I Progression, the first method of "strict" resolution:
