AbstractThe attempt to define living systems in terms of goal, purpose, function, etc. runs into serious conceptual difficulties. The theoretical biologists Humberto Maturana and Francisco Varela realized that any such attempt cannot capture what is distinctive about them: Their autonomy and unity. Goal, purpose, etc. always define the system in terms of something extrinsic, whereas living systems are unique because they maintain their unitary continuity of pattern despite the ceaseless turnover of their components. So, system‐closure is a prerequisite of their adequate conceptual comprehension. Maturana and Varela themselves found that system‐closure pertains exclusively to their organization, i.e. the set of relations among system‐components which unify them. For living systems this comprises the relation between the system‐components and the processes which they undergo. This relation is self‐referential because it is closed, i.e. it essentially (re)produces itself.While this model worked very well in the biological domain, attempts to extend it to the social domain met with serious conceptual obstacles. The reason for this is that Maturana did not make a consistent enough application of it. He understood the components of social systems biologically (individuals, persons, etc.) and the relations between them socially (language). This inconsistency ruptured the system's organizational closure. Consequently organizational closure (autopoiesis) can be maintained only when both the components of social systems and their processes are of the same type: Social. This interpretation can be found in the work of Niklas Luhmann who recognizes that the components of social systems are not persons, individuals, actors or subjects but communicative actions themselves. This preserves the organizational closure of the system and permits the concept of autopoiesis to be used as a powerful instrument of social