Luhmann N. (2000) Self-organization: Coding and programming. Chapter 5 in: Art as a social system. Translated by Eva M. Knodt. Stanford University Press, Stanford: 185–210. https://cepa.info/7858
Self-organization: Coding and programming.
Chapter 5 in: Art as a social system. Translated by Eva M. Knodt. Stanford University Press, Stanford: 185–210.
Fulltext at https://cepa.info/7858
Excerpt: We speak of self-organization whenever an operatively closed system uses its own operations to build structures that it can either reuse and change later on, or else dismiss and forget. Computers depend on external programming, although computer-generated programs may be developed eventually. By contrast, autopoietic systems produce their own structures and are capable of specifying their operations via these structures (structural determination). This mode of operation does not exclude causal environmental influences. Some of Munch’s paintings bear traces of water damage because they were left outdoors. While some people might consider this beautiful, no one would argue that the rain completed the painting. Nor would anyone try to prove the appropriateness of the rain’s decisions with regard to the altered formal structure of the painting. Rather, the impression is that a painting was not and could not have been painted in this manner. Self-organization owes its possibilities and its room for play to the differentiation of the system. Accordingly, art observes itself by means of the distinction between a reality “out there” and a fictional reality. The doubling of reality generates a medium of its own, in which the fixation of forms becomes not only possible but necessary, if the medium is to be reproduced. The opportunity and the need to do something go hand in hand. This conceptual model will guide the following analyses. In functional systems, we call the system’s basal structure – a structure that is produced and reproduced by the system’s operations – a code. In contrast to the concept of code in linguistics, we think here of a binary schematism that knows only two values and that excludes third values at the level of coding. A code must fulfill the following requirements: (1) it must correspond to the system’s function, which is to say, it must be able to translate the viewpoint of the function into a guiding difference; and (2) it must be complete in the sense of Spencer Brown’s definition, “Distinction is perfect continence,” rather than distinguishing just anything. The code must completely cover the functional domain for which the system is responsible. It must therefore (3) be selective with regard to the external world and (4) provide information within the system. (5) The code must be open to supplements (programs) that offer (and modify) criteria to determine which of the two code values is to be considered in any given case. (6) All of this is cast into the form of a preferential code, that is, into an asymmetrical form that requires a distinction between a positive and a negative value. The positive value can be used within the system; at the least, it promises a condensed probability of acceptance. The negative value serves as a value of reflection; it determines what kinds of program are most likely to fulfill the promise of meaning implied in the positive code value.