Addressing complexity in simulations using paraconsistent logic

Medical schools often use simulation as a technique for teaching surgical techniques. Some simulations are based on bi-valued logic - a binary relationship in computing terms, or a relationship based on simple probability. Here, particular levels of certainty are assumed for development of the simulation, allowing a simple framework for modelling basic scenarios. However, the real world represented in a simulation is unlikely to be bi-valued with two clear, definite solutions. Parameters will be fuzzy; they could be anywhere on a scale between these dual outcomes. Accepting this as a possibility, portrayal in a simulation becomes more complex and demanding. In surgical training, a simulation may need to take account of varying circumstances. It could be realized half-way through a procedure that the surgery is not really needed; or that the patient has unexpected physiological anomalies, so that a different approach is called for. Uncertainty brings more problems for modellers. Outcomes may call for subtler choices than a scale of values, e.g. ‘A or B, or possibly also C’ – modellers cannot be sure because the world is ambiguous. Many simulations fail to allow for situations such as this, which may give trainees a false sense of understanding created by over-simplification. Effectively, simulations based in bi-valued logic reinforce an illusion that there is always a clear solution, or that the solution is known. Human reasoning is not limited in this way. People are capable of assimilating ambiguity: keeping a range of options open or addressing complex issues that are summed up in the phrase ‘it depends’. Using paraconsistent logic, it is possible to envisage development of tools (including simulation software) that reflect human capacity to categorise and create resolutions that are inherently self-contradictory. This paper will set out a basis for simulation using paraconsistent logic that is tolerant of ambiguity.