According to [Francisco] Varela, an autonomous system can be precisely defined as a system that has organizational closure and operational closure (Varela 1979, pp. 55-60). The term ‘closure’ does not mean that the system is materially and energetically closed to the outside world (which of course is impossible). On the contrary, autonomous systems are thermodynamically far from equilibrium systems, which incessantly exchange matter and energy with their surroundings. ‘Organizational closure’ describes the self-referential (circular and recursive) network of relations that defines the system as a unity. At any given instant or moment, this self-referential network must be maintained, otherwise the system is no longer autonomous and no longer viable in whatever domain it exists. ‘Operational closure’ describes the recursive, re-entrant, and recurrent dynamics of the system. The system changes state on the basis of its self-organizing dynamics (in coupling with an environment), and the product of its activity is always further self-organized activity within the system (unless its operational closure is disrupted and it disintegrates).7 Biological examples abound—single cells, microbial communities, nervous systems, immune systems, multicellular organisms, ecosystems, and so on. Such systems need to be seen as sources of their own activity, and as specifying their own informational or cognitive domains, not as transducers or functions for converting input instructions into output products. In other words, the autonomous nature of these systems needs to be recognized.
Neurophenomenology: An Introduction for Neurophilosophers (pdf)
Evan Thompson, Antoine Lutz, and Diego Cosmelli