Many early studies of organisations tend to use the language of variables in the form of so-called trivial input–output models (‘trivial’ is their official name, not the author’s; von Foerster 1993). Over time, the models used have become more complex and have been expanded to include parameters (context-dependent variables) and state-spaces (Chernoff 1953).

A second development saw the use of control models; models that are used to effectively achieve or maintain some aim or objective (Beer 1966). A third group contains models of organised collectives, i.e. models that identify how a number of elements (not necessarily with human-like behaviour) are or become linked (Kaufman and Rousseauw 1990). A fourth type refers to coordination systems or groups of elements the aim of which is to perform some activity (Hollan 1992). The four types (and the studies in which the models have been used) identify an inability: each new type of model has been introduced to deal with the problem of not being able to achieve research quality (in the sense of Comte 1854, and others; see above). In other words, none of the models appears to provide the ‘twofold’ name being looked for—leaving the naming of the data as their best contribution. The main difficulty seems to be that the models depend on observation, whilst organisations strongly depend on objectives, plans, visions and the like.