The distinction between being and nonbeing is now replaced as primary distinction, completely implausible from an ontological point of view, by the distinction between inside and outside or that between self-reference and other-reference. For according to the new version, an observer has first to be produced before he can apply the being/nonbeing distinction. (Luhmann, Theory of Society, vol 2, 195).
The distinction between system and environment and the re-entry that produces self-reference and other–reference must precede the observation of being. For present purposes, let’s think of observing systems. How is an observing system produced? By drawing a system/environment distinction. This distinction allows the distinction between self-reference and other-reference, which, then, allows for observation.
Back to Luhmann:
The prevailing view of the world in old Europe can be described as ontology. . . . [The] predominance of the ontological mode of observation is demonstrated alone by the fact that paradox was invented to defend Eleatic ontology and thus treated from the outset as a disturbance if not error in reasoning to be avoided; and, moreover, that bivalent logic, which ontology had relied on to block reflection, was unquestionable accepted until very recently.
By ontology, I mean the result of a mode of observation that operates on the basis of a distinction between being and nonbeing and that subordinates all other distinctions to this one. This distinction finds its inimitable plausibility in the assumption that only being is, and that nonbeing is not. This is then taken over in logic as the law of the excluded middle [tertium non datur]. . .
For present purposes, ontology is defined as an observation schema that orients itself on the distinction between being and nonbeing, and thus independently of the very different ways in which philosophy has treated the concept. This means above all that the distinction between being and nonbeing is and remains dependent on a primary operational separation, namely, of observation (or observer) from what is observed. (Theory of Society, vol 2, 185)
We must be careful not to equate this usage of term observer with the subject of the discredited subject/object worldview. This is why I want to think in terms of observing systems, which are not subjects.
Luhmann is arguing that the system/environment distinction precedes or allows the being/nonbeing distinction because the observing system must first be produced (by itself!) before any observation can happen. A system produces and reproduces itself by marking a border between itself and its environment–which is everything else. Observation cannot happen without a distinction because “everything” cannot be observed at once, which means that an inclusion/exclusion schema is needed. Observation relies on a “blind spot” because everything cannot be observed at the same time.
Classical ontology portrays the world as a whole that consists of parts–and the only thing indivisible is being. Moreover, each part has a designated, natural place–and only one place. Systems theory, in contrast, takes difference as its point of departure.