Listen by Control-clicking on the link directly beneath and selecting "Open in New Window"

or

Download for free by Control-clicking and selecting "Download linked file" or whatever the Windows version of these things are

This is the link you should do those things with: Hear The Music You Are About To Read About

(It will take a while. Wait for it. It is not a three minute pop song, as you will soon see.)

I bought a two disc set from the Detroit Symphony Orchestra called Black Composers Series. It had some 12 tone inspired stuff on it. It made me want to write a very strict serialist piece. After listening to it for the first time on the night of Tuesday, May 21, 2002, I made a note to myself to do it the next day, but then started it that night instead, even though it was late.

I wrote the alphabet. I wrote the chromatic scale directly underneath the alphabet, one note to one letter. A and Bb were included three times, while all other pitches were included twice (26/12 = 2, remainder 2) I wrote my girlfriend's full name. I wrote the corresponding notes underneath the letters of her name, then rejected duplicate notes. This resulted in the following row:

G F A Bb C# D G# B C E

Not all 12 notes were included, so I thought I might try the cats' names. One added Eb, another added F#.

I decided I couldn't decide which order the resulting notes should go in. I chose my own full name instead, which added, in this order, F# Eb

This was my resultant home row, row one.
G F A Bb C# D G# B C E F# Eb

I flipped several coins, one at a time, to determine whether each note would be higher or lower than the preceding note. I had quarters from MD, NY, CT, VA, PA (2), MA, and GA, all minted in Denver, a 2000 Sacajaweja dollar minted in, I believe, Philadelphia, and a battered 1994 penny with no readily apparent information on where it was minted. The first flip was heads, so I drew an upwards pointing arrow between the G and the F. I continued to flip the coins one after another until there was a directional arrow between each pair of notes. Now I had the pitches for row one.

To determine how long each note would last, I chose six of the state quarters - PA, MA, VA, NY, CT, GA, and assigned each of them two pitches, one for heads and one for tails. I shook, shuffled, and cut all six quarters in my hand, thoroughly randomizing them. Then I stacked them and fanned them out in a horizontal line. The first shake, shuffle, and cut of the six quarters came out like this:

PA MA CT NY GA VA

The first coin in the first line was PA. This represented either the note A (heads) or the note Bb (tails). This note would get .5 seconds or beats. (I decided on a tempo of 60 beats per minute.) I flipped the MD quarter to determine whether I would use the heads note or the tails note.

In this chart each state is followed by its heads note then its tails note.
PA A Bb
MA G# B
VA G C
NY F# C#
CT F D
GA E Eb

The MD quarter came up heads, so the note A got .5 seconds written underneath it. The second flip of the MD quarter was tails, so the MA note was B. B got 1.0 written underneath it.

When in this manner I had established which notes would get .5, 1.0, 1.5, 2.0, 2.5, and 3.0 seconds, I shook, shuffled, and cut the 6 state quarters again. This time they represented whichever note of theirs they did not represent the first time. Using the resulting horizontal line, I determined which notes would get 3.5, 4.0, 4.5, 5.0, 5.5, and 6.0 seconds.

Now row one had pitch and rhythm. For both pitch and rhythm, I applied exactly the same approach for rows two through six. For the seventh row and all the ensuing rows, I made a slight modification. I allowed the side of the coin that landed up in the horizontal line of six state quarters to determine which note would receive the time allotted. I feel this was slightly more elegant than using the MD state quarter. In a way, it represented my family leaving that fine state. Also, the back of my left hand was getting sore from all the slapping of the MD quarter.

Once I had pitch and rhythm for each of the twelve notes in each of my twelve rows, I sketched them out without rhythm to look at the ranges required to play them. The largest range was three octaves and an Augmented fourth (42 half steps). The smallest was one octave and a Major third (16 half steps). I assigned appropriate instruments to each of the twelve rows. (I had already somewhat hastily decided that the first row would be played by animating the frequency of a signal generator on ProTools and that the second row would be played on the electric piano. Luckily, the ranges worked out very well for both.) I finished this and began recording the rows on Wednesday and Thursday.

Now I have to decide when to play each row. I don't think end to end will be very interesting, but it would be the purest serialism. Perhaps I will start the piece with them all end to end - row one, row two, etc. Then I may find some way of randomizing their later appearances and allowing them to be played together while maintaining mathematical, aleatoric, serialist purity and avoiding tiresome traditional harmony. This may end up being a very long piece.

5-30-02
I figured it out last Thursday. The larger structure of the piece now consists of 11 parts.
Part 1 is rows 1-12 in order.
Part 2 is random rows 2 at a time
Part 3 is random rows 3 at a time
Part 4 is random rows 4 at a time
Part 5 is random rows 6 at a time
Part 6 is all 12 rows at once
Part 7 is random rows 6 at a time
Part 8 is random rows 4 at a time
Part 9 is random rows 3 at a time
Part 10 is random rows 2 at a time
Part 11 is rows 12-1 in order.

I made this chart, assigning a certain coin result to each row. The state name is followed by its heads result, then its tails result.
PA 1 9
MA 2 10
VA 3 11
GA 4 12
CT 5 7
NY 6 8

I made this chart to be filled in:

Part 1 is rows 1-12 in order.

Part 2 is random rows 2 at a time
Pair 1:
Pair 2:
Pair 3:
Pair 4:
Pair 5:
Pair 6:

Part 3 is random rows 3 at a time
Trio 1:
Trio 2:
Trio 3:
Trio 4:

Part 4 is random rows 4 at a time
Quartet 1:
Quartet 2:
Quartet 3:

Part 5 is random rows 6 at a time
Sextet 1:
Sextet 2:

Part 6 is all 12 rows at once

Part 7 is random rows 6 at a time
Sextet 1:
Sextet 2:

Part 8 is random rows 4 at a time
Quartet 1:
Quartet 2:
Quartet 3:

Part 9 is random rows 3 at a time
Trio 1:
Trio 2:
Trio 3:
Trio 4:

Part 10 is random rows 2 at a time
Pair 1:
Pair 2:
Pair 3:
Pair 4:
Pair 5:
Pair 6:

Part 11 is rows 12-1 in order.

Then I began shaking, shuffling, and cutting the six state quarters while looking away from my hands and then slapping one down on the table. The first one was the NY quarter with the tail side up, so I wrote the number 8 right after the line "Pair 1:" in Part 2 of the piece. Then I put the quarter back in with the other five and did it over again. Once a given side of a coin had come up once (that is, a row had already appeared in a Part of the piece), that row could only appear in the following Part. I followed this strictly.

Even once it was obvious, by the process of elimination, which rows would appear in the final Pair, Trio, Quartet, or Sextet of a Part, I waited for them to appear naturally before writing them in. This took somewhat longer than it would have otherwise, but I wanted to let the coins have their full say. However, once a coin had displayed both its heads and its tails sides at least eight times each (that is, put both rows it could represent in all 8 randomized parts of the piece), I removed it from the pack. I would estimate that it took about 120-130 shakes of the coins.

The quarters were removed from the pack in this order: PA GA VA NY MA CT. (This fact can be deduced from studying the finished Part 10, below.) I find it interesting and possibly significant that the PA quarter's tail side came up at least six extra times after row 9 had been assigned but before row 1 had been assigned. September will always be the month that I met my girlfriend. Here is how the page looked when I was done:

Part 1 is rows 1-12 in order.

Part 2 is random rows 2 at a time
Pair 1: 8 9
Pair 2: 11 1
Pair 3: 4 10
Pair 4: 12 6
Pair 5: 5 2
Pair 6: 7 3

Part 3 is random rows 3 at a time
Trio 1: 9 1 11
Trio 2: 6 12 10
Trio 3: 4 5 7
Trio 4: 3 2 8

Part 4 is random rows 4 at a time
Quartet 1: 11 9 6 1
Quartet 2: 12 10 4 5
Quartet 3: 7 3 2 8

Part 5 is random rows 6 at a time
Sextet 1: 9 12 1 4 10 6
Sextet 2: 5 7 3 11 2 8

Part 6 is all 12 rows at once

Part 7 is random rows 6 at a time
Sextet 1: 9 12 4 6 1 3
Sextet 2: 7 10 11 5 8 2

Part 8 is random rows 4 at a time
Quartet 1: 12 9 4 1
Quartet 2: 6 3 7 10
Quartet 3: 11 8 2 5

Part 9 is random rows 3 at a time
Trio 1: 9 4 6
Trio 2: 3 1 10
Trio 3: 7 12 11
Trio 4: 2 8 5

Part 10 is random rows 2 at a time
Pair 1: 9 4
Pair 2: 6 3
Pair 3: 10 1
Pair 4: 12 7
Pair 5: 11 8
Pair 6: 2 5

Part 11 is rows 12-1 in order.

With each row lasting exactly 39 seconds, the total length of the piece will be 39(12+6+4+3+2+1+2+3+4+6+12)=2145, or 35 minutes and 45 seconds.

I find the fact that rows 2 and 5 appear together in both Part 2 and Part 10 interesting, especially since they also appear together in Parts 5,(6),7,8, and 9. But this fact, while interesting and very possibly significant, is also possibly bordering on minutiae.

I am eager to finish the piece and see how it sounds. I am stalled on recording it right now while I finish and archive some other work, then defragment my hard drive. My computer started acting a little buggy while I was working so I decided to do this.

I realize that there could be a tremendous variety of pieces written with exactly this same process. It would be possible, though not necessary, to use different names to determine the home row. My mom suggested that maybe when Kim hears this piece Kim will realize something important about the very nature of the universe, since it was based on her own name. That's just one reason I love my mom. It's possible.

2-12-03
Well, after several incidents pleasant and unpleasant (wedding, catastrophic hard drive failure, etc.) stalled work on Georgia Tails, Connecticut Heads, version 1.0 is finally playing on my CD player. It is actually fairly pleasant and inobtrusive to have on while doing things around the house. In performance however, it might be something only the most dedicated "new music" people would sit through. I might replace some of the more "homey" instruments (casio keyboards, my own trumpet playing) with other instruments or at least better players if it were going to actually be performed in real time. But I might not. It could be something to meditate to. It would be great for that. I like it a lot. There are some really nice sounds.

I have yet to figure out why the resulting piece was not exactly 35 minutes, 45 seconds. This puzzles me but I don't have time or necessary resources and information right here at this moment.

Wayne Everett is a founding member of the two man dicsco collective little blind fireman also featuring Earl Alexakiss.

Update: The resulting piece IS exactly 35 minutes, 45 seconds. I don't know why I thought it wasn't.

Prose Index
Band Index
Memphis Evans Home