Someone suggested widecasting as an alternative to sidecasting yesterday so I am trying it out in the title here. Not sure myself 'side' has a wicked element to it that 'wide' does not provide but nothing ventured nothing gained as they say. Yesterday I referenced Dreborg's 1996 article on backcasting and its use of a table to contrast forecasting with backcasting. I found that useful so I have replicated it below, but with an additional column for sidecasting. My plan is to summarise that table today, then after a break I will move onto to future backwards as a technique along with others. In the advanced courses in Melbourne and Auckland I am going to be teaching some of the new material on coherence and consensus as well providing a flow chart of these various methods so there is a short term opportunity for people interested in taking future backwards (easily one of our most popular methods) to a new level.
Aside: I haven't yet managed to get tables sorted out in Expression Engine (the tool we use for this web site) so I've cut and past an image instead and put it at the bottom of this post. Apologies if its a bit over-wealming or there are problems with rendering.
For Forecasting and Backcasting I have replicated Dreborg's words and they speak for themselves. Hopefully my last two posts have properly set the context. So lets run through the first two of his headings (I will reserve the last three for a future post)
Philosophical views differ to a degree between forecasting and backcasting, the latter being teleological, focusing on final causes or more pragmatically end states. The question of evidence moves from justification by mathematical extrapolation from the present state to discovery and a recognition of indeterminacy. However both of the methods are linear (they differ in direction only) and they also assume causality. Here we get the biggest difference; a complex adaptive system is dispositional but its not causal in the conventional sense. That means we can map the dispositions of the system to evolve in some ways and not evolve in others. However we do not get the repetition of outcome that allows us to use causality in any meaningful way: the same thing will only happen again the same way by accident. We also shift from thinking about drivers (common to most foresight approaches) to modulators.
Perspectives again show a similar pattern with a significant difference between the quasi-determinism of forecasting and the more choice based backcasting. Sidecasting shares more with backcasting here, but again is different. It is taking an evolutionary not a design approach to the issue. It seeks to test the evolutionary potential of the present before determining direction. You might choose a difficult path which goes against the flow of events, but you need to be aware of the energy cost and risk of such an approach. As any hiker knows the shortest way between two points is rarely as the crow flies. Critically we also seek to be highly adaptive, managing the evolutionary process to enable us to discover solutions and path ways which could not have been predicted in advance (forecasting) or known (backcasting), but which are more sustainable and resilient.
I will return to the other three in my next post on this subject but in the meantime, here is the table: