Methods in futures studies: A view from the theory of science

Abstract Viewing forecasting in terms of three modes, as predictions deduced from tested scientific theories, as trend extrapolations, and as heuristics, it is shown that most long-term predictions are inherently contingent. They do not tell us what will happen unconditionally, but what will happen, in case something else happens. Hence futures studies will not show us any predetermined future, but rather different ways of shaping the future. Scientific predictions, trend extrapolation, and heuristic methods are the most important approaches to forecasting. If they are properly understood, this will help understanding other less important methods also.