Notebook

Notebook, 1993-

RELATIONSHIPS

Autocorrelation of Form









. . . . But what is form? The theorists define it thus: a group of elements perceived in their totality, as it were, and not as the product of any chance assemblage. [I emphasize "perceived," for of course form in reality may well be the result of a chance assemblage.] To create a form thus defined would mean to assume a certain previsibility of this form. In the strictest sense of the word, previsibility, or Vorsicht, means to imagine a phenomenon in the future, in terms of the past. Let us imagine a circle we watch someone drawing. At any moment the designer can interrupt the design, while for my part I can at any moment hazard a guess as to what he intends it to be. If, at the instant he stops drawing, the circle was almost closed, I can be the surer that he wants to draw a circle than I can when he is just beginning to draw. Thus there obviously exists a degree of previsibility, a connection that is at least a statistical one between past and future, a correlation between what happened just now and what is going to happen in the immediate future.

It is this connection between the past and the future of a form explained statistically that Wiener has called "autocorrelation." Autocorrelation evidently varies between 0 and 1. It is 0 when the phenomenon is totally lacking in order and its behavior in future is therefore not previsible. As order appears, it tends toward the value of 1, or the autocorrelation of a completely ordered phenomenon--in other words, one that is indefinitely previsible. At first sight, we see that autocorrelation expresses very well the law of good form in the Gestalt theory. The nearer autocorrelation comes to 1, the better the form. However, autocorrelation differs in one essential from good form, since the later does not separate into parts.

[The Problem of Form p. 209. Molnar, Fran┘ois. "The Unit and The Whole: Fundamental Problem of the Plastic Arts." In Module, Proportion, Symmetry, Rhythm. Vision and Value series. Gyorgy Kepes, ed. New York: George Braziller, 1966.]




NOTEBOOK | Links

Copyright

The contents of this site, including all images and text, are for personal, educational, non-commercial use only. The contents of this site may not be reproduced in any form without proper reference to Text, Author, Publisher, and Date of Publication [and page #s when suitable].