So finally the summary to my set of posts about the validation and interpretation of Socio-Ecological Simulation Models (SESMs)that arose out of some of the thinking I did during my PhD thesis.
The nature of open systems requires SESMs to specify and place boundaries on the system such that it may analysed effectively. Recent debate in the geographical and environmental modelling communities has highlighted the importance of observer dependencies when identifying the appropriate model ‘closure’. Furthermore, because an ‘open’ system can be ‘closed’ for study in multiple ways whilst still adequately representing system behaviour, the issue of model ‘affirming the consequent’ is present when attempting to model these systems.
Because of these issues I suggested that a more reflexive approach, emphasising trust via practical adequacy over the establishment of true model structure via mimetic accuracy, will put SESMs in a better position to provide understanding for non-modellers and contribute more readily to the decisions and debates regarding contemporary problems facing many real world environmental systems.
This is not to say issues regarding mimetic accuracy and model structure should be totally ignored – these model validation criteria will still have a role to play. However, emphasising trust via practical adequacy over truth via mimetic accuracy, ensures the model validation question is ‘how good it this model for my purposes?’ and not ‘is this model true?’. Engagement with local stakeholders throughout the modelling processes, contributing to model development and application should ensure practical adequacy, but also, in parallel, trust. As a result of this participatory model evaluation exercise, confidence in the model should be built, hopefully to the level where it can be deemed to be ‘validated’ (i.e. fit for purpose).