there is a difference between regulation and theorizing. regulation pertains to efficient action, in ashby's terms: counteracting disturbances and the law of requisite variety holds there. theorizing pertains to a simplification of the theorized (object or system) which represents some of its stable features. a theory must be true or at least validatable by evidence. all simplification amounts to a loss of variety. the natural science deliberately loose variety. klaus ________________________________ Från: PhD-Design - This list is for discussion of PhD studies and related research in Design genom Mattias Arvola Skickat: sö 2008-10-12 05:56 Till: [log in to unmask] Ämne: Re: information as an entity rather than an activity On Fri, 10 Oct 2008 15:31:13 -0400, Klaus Krippendorff <[log in to unmask]> wrote: >you are saying that a theory must have more variety than what it represents. just the opposite is the case. any theory reduces the variety of phenomena to the variables it theorizes. the theory of free fall does not say anything about aerodynamics and weather conditions of the experiments conducted. Perhaps I can claryfy what Ashby meant by the law of requisite variety. For a regulator to be able to control or regulate another system it needs to have a model of that system/process that has more possible variety than the system/process it is set to regulate. The reason is that the regulator, to be able to control the process, it needs to be able to meet every disturbance with at least one counteraction. Thus it needs to have more variety than the process it is supposed to control. How we should translate this to theories I'm not sure. But I would expect that if we want to explain or predict a phenomenon, we would need our theory to be able to predict at least all states of the phenomenon: it would need as much or more possible variety than the phenomenon it is supposed to explain or predict. However, my memory and interpretation of Prof. Erik Hollnagels cognitive system engineering classes may be faltering. Cheers, // Matti Arvola