The problem of comparison in soft systems methodology

The core idea in soft systems methodology is to generate logical systems models that are thought to be relevant to a problem situation and then to compare the model and situation in order to structure a debate about change. It is demonstrated in this paper that comparison is problematic both in theory and practice. A learning-based stream of research is described in which three concepts, termed expectation, desirability and importance (EDI), are identified to help explain the nature of comparison. The learning gained from applying these concepts in order to structure practice is discussed. Copyright (C) 1999 John Wiley Rr Sons, Ltd.