Fantasy in F

I began work on this piece in 1979, when I was living in San Francisco. Describing its origins, I later wrote
"I was hoping to capture the feeling of the city's buoyancy, optimism (mine was sorely lacking at the time;
I was too poor to feed myself adequately, my health was suffering, and my girlfriend had just thrown me out),
and hills in the left hand's opening ostinato figure."

I don't remember how much of the piece I'd written by the time I moved back to my home town, Santa Barbara, later that year, but I didn't complete it until 1981, in time for pianist William Koseluk to perform it at a concert composer John Carbon and I presented (poster, linoleum block).

A decade later, I sent a copy of the score to the San Francisco Symphony's pianist Robin Sutherland. He phoned me to say that he'd gotten it, and that he liked the piece. Later, in response to my description of how I came to compose it, he sent this letter, which shined light on my hopes for the piece for a long time.

He never did get around to performing it, though, and over time I got accustomed to the likely reality that he probably never would.

After that, I didn't think much about the piece until I was a few years into the next adventure: posting my animated graphical score videos on YouTube. In 2008, after I'd done a few other videos based on my own compositions, I published a simple "piano-roll" style video of the Fantasy.

The idea to do a version using harmonic coloring was so obvious and logical that I actually thought I'd already done it, and was surprised to find (in June of 2021) that I hadn't—so I did. This video is the first version I published (and its FAQ contains links to other versions I made later). Here's the latest version:

I was a little disappointed that while that video made the quartal structure visible in individual, local elements (chords and arpeggios), the tonal/harmonic motion was less easy to see. I made this picture of its harmonic motion (pitches colored and ordered vertically according to the circle of fifths) ...

... but while that did give a clear overview of the motion (fast at the beginning and end, slow in the canon and fugue straddling a fast center section), there was still something missing.

As part of my work with Easley Blackwood's microtonal etudes, I developed a display that showed pitch motion within the circle of fifths, and this turned out to be just what the Fantasy in F needed:

Watching this, I realized that in the forty years since I wrote it, I'd completely forgotten how it worked. What I saw was this: over and over, things start out well-behaved, with notes staying within small regions in the circle of fifths, but after a bit they start moving further afield, until finally, it goes crazy, jumping all over the place (nowhere near where it started), and then has a cadence (not a conventional triadic V → I, but something like it) back to the starting point.

The trick to going between well-behaved to wacky is to play with the two different scale step sizes. The major second (whole step) moves you the same distance in the circle of fifths as two steps by fifths or fourths—not far at all—while the minor second (half step) is equivalent to five fourth/fifth steps—nearly halfway around the circle of fifth (and firmly into "we're not in Kansas anymore" territory). This means that chromatic motion (like at the end of ostinato) is pretty much the maximum in chaos before returning home.


The piece is a single movement, but it's divided into several fairly clear sections (measure numbers for this version of the score):

  • Passacaglia (though just barely)  The first eight measures present the seed of the piece (in F), a bass line built from arpeggios of fourths and fifths. This seed is then repeated two times with free treble accompaniment. The opening harmonic move of the seed begins a passage (mm 25) with the first semi-melodic material in the treble, subsequently echoed in the bass), leading to a rising canon (mm36) which concludes with a cadence in B-flat. A short version of the opening material in that key (mm43) is followed by a cadential modulation to E.
  • Relaxed  This section (mm 52) explores the material (fourths and fifths) in a slower tempo, modulating slowly back to the home key (mm 77).
  • Canon  The notable feature of this two-part canon (mm 78) is that while the main rhythmic patterns of the canonic line are in 3/8, the canon is at the distance of a half-note. The result of this is that it's harder to anticipate the movement of the following part (comes). The canon stays around the F tonality for a while, then modulates back to E.
  • Relaxed  (mm 129) Almost the same as the first time, but with a couple of things to make it more intense: some wider intervals in the bass, and more rhythmic compression at the end (leading to a cadence in D-flat—the last important note in the passacaglia).
  • Fugue  Except for the rhythmic irregularity and quartal harmony, a fairly conventional fugue. A notable feature is that it morphs back into the "relaxed" material (mm 186). Like the canon, it modulates slowly (to D, mm 204).
  • Dream of Passacaglia  At a much slower tempo, fragments of the passacaglia theme and other material from the first section are used freely, leading to a truncated version of the rising canon (mm 231).
  • Passacaglia  The cadence at the end of the Dream is a little misleading. The final chord (end of mm 234) feels like a dominant (V) on F, which makes the return of the opening passacaglia (this time in octaves) feel at first like it's in B-flat, but by the end, it's clear we're back in F. This time through, though, the semi-melodic passage is a step higher, which allows the rising canon (mm 268) to be on C instead of F, and thus cadence on F, to set the stage for the coda.
  • Coda  If you remove the flywheel escapement from a pocket watch, there's nothing to regulate the speed at which the hands turn, and they spin around like crazy. That's what the coda reminds me of.

    I should probably say a little more about a notable characteristic of this piece: its irregular rhythms. The rhythm of each line alternates between eighth notes and quarter notes, and the rhythms of the lines (usually two, but sometimes three in the fugue) are seldom aligned with each other. In the original score of the piece (which looked more or less like this), I didn't use barlines (since they often conflicted with one line or the other), but this turned out to make it much harder to read (since there were no "anchor points" to hang on to), so I chopped it up, but this is only a notational convenience; the way to approach this piece as a performer is to take each line separately, and figure out its accents, phrasing, and shape on its own.


    At this point there are several versions of the score floating around. They all have more or less the same notes, but they vary substantially in other ways: layout, barlines, dynamics, articulation, accents, fingerings, etc. I've also taken a stab at reworking it for string quartet (not as much a stretch as you might think; I conceived of the piano version as a kind of "string quartet for keyboard," and only went outside the range of string instruments for one note — an octave doubling on the final chord). This ZIP archive has everything, I think (including a README that says what's what). For a quick glance: score

    Stephen Malinowski — July 3-11, 2021
    Please contact me with questions, corrections, or comments.