Monday, January 11, 2010

Face / Light / Leaf

Until the face exposed them, the brain did not detect its
features. On paper, the half-face hybrids darkened: Blind
face impression, separate from the passive nameplates.

Until the light exposed them, the brain did not detect its
features. On paper, the half-light hybrids darkened: Blind
light impression, separate from the passive nameplates.

Until the leaf exposed them, the brain did not detect its
features. On paper, the half-leaf hybrids darkened: Blind
leaf impression, separate from the passive nameplates.

-- Glenn R. Frantz

Croatia's break with the past
Even in blind patients, light worsens migraines
Even as electric cars are rolled out in Detroit today, U.S. will lag the world in adoption

1 comment:

Anonymous said...

I used my usual toolkit of custom text-randomizing programs, manually selecting and assembling the choice bits of output. The difference was, to the input I added a second copy of the source text, in which I had manually replaced most of the nouns with a placeholder symbol ('@'). Then when I read the output, I imagined the placeholder replaced by each of the three subject nouns (which I selected in advance), to find phrases that worked with all the subjects. For example, "Until the @ exposed them".

In this case, the stable part is so short that it was almost all tweaked by hand by the time I was done. But I still used the randomized text to generate the ideas.

Technically, I should be able to dispense with the whole placeholder technique, just pick out one of the nouns in a phrase of regular output, and imagine replacing that noun with each of the subjects. But I find that the presence of specific words everywhere is too distracting for me to be able to do that! I need to have "blanks" to fill in.

I've previously used this technique to write "duo" poems; for example "Bicycle / Island":

http://stridemagazine.co.uk/Stride%20mag2009/july%202009/frantz%20july09.htm

So I figured this trio form would be appropriate for 3by3by3. But after writing a couple of them, although I think the results are worth reading, I'm coming to the conclusion that two repetitions is optimal and three is just a bit too much! So if I do something like this again for 3by3by3, I'll have to find some way to tinker with the pattern.

-- Glenn