Dealing with the harmonic boredom issue

I delved into my extensive library (no exaggeration- we have 48 overflowing bookcases at home.) again today. The impetus behind today's canon comes from Fourteen Canonic Inventions by Meyer Kupferman [1926-2003] (NY: General Music Publishing, 1966).  It's a useful and interesting volume in which the composer, himself, analyzes the compositions that he offers as models of canonic technique.

The second canon in the book is called "Canon on a Choral."  [Out of respect for the copyright owner, I will not reproduce it here.]  Kupferman wrote a one-measure chorale phrase based on the harmony i, iv, V in the key of A minor.  He then took the four parts of the chorale "linked together in a linear fashion" to create the 4 measure melody.

It's an interesting technique to generate a canon but it does have its limitations.  The most obvious of these is that the harmony will repeat for the duration of the canon.  This is a perennial problem for many types of canon.

The task that I set myself was to use this particular technique but to try to overcome the harmonic boredom issue, at least for this type of small form.

I wrote a two-measure chorale phrase with the harmony i - III - N6 - i6/4 - V.

I chose the Neapolitan chord because of its function (pre-dominant) and because of its pitch content, as you will see below.  The Neapolitan chord is a major chord built on the flatted second degree of the scale (bII) -- in the key of E minor it is an F major chord.

By making the rest of the chorale diatonic and saving the F natural for the end of the melody, I was able to delay the chromaticism and the piquant Neapolitan chord until late in the piece despite the repetitious nature of the canon.

The first beat of each of the even-numbered measures has the pre-dominant function. In measure 6, that function is handled by only two pitches ("A" and "C") -- ambiguous, perhaps --but in the diatonic environment, we'll likely hear it as iv.  In measures 8 & 10, the added F natural makes that pre-dominant function a very unambiguous Neapolitan.

To add interest, I allowed myself the license of a change in the bass melody in the antepenultimate measure.  This half note "C" creates a deceptive cadence which adds just enough interest to allow one more repetition of the two-measure "chorale" which brings the canon to a close on an authentic cadence.  This simple (and common) rhetorical device also, in this case, brings the violin around to the beginning of the canon again to recap the first two-measure phrase.

I've included an outline of my procedure and a harmonic analysis.  Compare, if you will, the analysis of the chorale as I first wrote it and the analysis of the completed canon.

(click on image to enlarge)