Niklas LUHMANN is well known for his deliberate departure from the classical focus on studying individual actions and directing attention on the actors' relatedness through so called (autopoietic) communication systems. In contrast to the gain of a new perspective of observation his focus on autopoietic systems is simultaneously its biggest methodological obstacle for the use in social and management sciences. The present contribution considers the above shift on a theoretical level and with a specific qualitative method. It argues for a deeper understanding of systemic sense making and its enactment in a systematic and comprehensible way. Central to this approach is its focus on groups. Using group discussions as the method of data collection, and the "documentary method" by Ralf BOHNSACK (2003) as a method of data analysis, the article describes a methodologically grounded way to record the self-referential systems proposed by LUHMANN's system theory. The theoretical considerations of the paper are illustrated by empirical vignettes derived from a research project conducted in Switzerland concerning the social responsibility of business. URN: urn:nbn:de:0114-fqs1003156