On Method -- Mary Margaret Sloan

"On Method" began as a reading of Descartes' Discourse on Method. I wanted the poem to involve a method that would explore a classical model of order with a mathematical basis and would then wreck itself from within, deteriorate that classical symmetry and develop as a complex adaptive system into a more flexible model of order. And while I was interested in finding a way to open out an orderly form, at the same time I was looking for a way of returning to something traditionally strict. To escape the boredom of free forms or forms organized as metaphorical models, I wanted a procedure that would permit exploration of the pleasures of restraint, in particular, of mathematical restraint.

Right away I discovered a difficulty in creating a mathematically based sequence of change gradual enough to be discernible to the reader; if the size of the unit of repetition (anything from a letter on up to stanza) as well as the increments of change were too great or too small, the alteration would be experienced by the reader not as an orderly pattern of change but as an entirely random - and therefore free - form, not the sense I wanted. Initially I tried using a fractal model but couldn't discover a repeatable prosodic unit that would lend itself to a suitable rate of growth based on a fractal ratio; in relation to the countable elements of a poem (syllables, words per line, lines, and so on), the rate of growth was too rapid to be felt.

I finally decided to use a modified version of a rondel as the basic form because of its thirteen lines and orderly sequence of repetitions with the central line of each poem recurring in the next poem in a fixed order. This form had the advantage of the odd number of lines and thus a central point with its suggestion of classical axially ordered symmetry; this balanced model could then be subjected to a process of mathematically determined deterioration which would throw it off center and result in an order of regular irregularity. At the same time, the procedure would have the effect of causing the poem as a poem to fall apart by gradually extending some line lengths with each successive repetition further towards the right hand margin until eventually the poetic line would disappear, that is, would run over and gradually turn into prose.

To explore the symmetrical order of the rondel, I first ran the poem through thirteen repetitions with the central line of the first poem repeated as the last line of the second poem, then the central line of that second poem repeated as the first line of the third poem, next as the penultimate line and so on back towards the center.

After completing the first cycle of repetitions, I started over again, reiterating the formula of repeating the central line of the first poem in the next, but this time moved the line to a random location and made the character of the line gradually transform. As the number of words in the repeated line changes (first diminishing, then gradually increasing), the words themselves of the original line decrease by one each time the line is repeated and each time are rearranged and mixed in with new words that appear in order to make up the new line's requisite number. In this way, the lines acquire an identity (the line with a particular set of words) which is then lost as the original words of the line gradually wash away, or are diluted, as the line reappears in each successive poem until all that remains of it is its ever-expanding place which at some indeterminate point mutates from a line of poetry into a unit of prose.

I still needed a mathematical device to generate an orderly increase in the number of words per line. Since I couldn't find a way to apply the fractal ratio (and maybe somebody else could - I'd love to know), I turned to quadratic functions [f(x) = ax2 + bx + c] which seemed appropriate for the role they play in Cartesian analytic geometry and in particular for the manner in which, when graphed, they display themselves around a central Cartesian axis. They were also nicely malleable as generators of increase. I tried various equations until I came up with one in which the independent variable, used to establish the number of words in successive repetitions, would increase at a perceptible rate.* In order to make the model change in other interesting ways , I introduced a few arbitrary elements having to do with positioning of repeated lines as well as a ratio determining the number of lines per poem and the number of words and lines per stanza. The cycle, as of now, has twenty-two poems but may be continued later on.

* The form of the equation is: y = ax2 - bx + c. x = the number in the order of repetitions of a line (x = 1 for the first time it's repeated, x = 2 the second time, etc.), a = 1, b = 2, c = the number of words in the line the first time it appears (so c = 7 for the first line that's repeated because the line originally has 7 words), y = the number of words for the line in the next poem of the cycle. Therefore the equation for the first repetition when x = 1 and c = 7 was y = x2 + 2x - 7, or y = 1 - 2 + 7, or y = 6. In the second repetition, y = 7, in the third, y = 10, and in the last poem, y = 55.