Dynamic Order: The Problem of Method in Evolving Nature : With Letters from N. Rescher, L. Pauling, J. Eccles, and K.R. Popper (Millennium (Series), 6.)

At one time, scientific method seemed to satisfy the need for universality and necessity on which modern science was based. It was possible for everyone to retrace the same methodological itinerary, with the assurance of reaching certain results. Today, owing to the complexity of reality, this conviction seems to be no longer tenable. But it would be a great mistake, nevertheless, to maintain that it is pointless to talk about method.
Even disorder itself, when disassembled according to a criterion, shows some possibilities of order, and thanks to the help of variables, can be understood and represented graphically. We find ourselves faced with a paradox: from apparent casuality, order begins to emerge little by little - certainly not a static order of the Aristotelian or scholastic type, but a dynamic order which, once a certain threshold is passed, runs the risk of hurling itself back into chaos, and, often, of preparing once again for some forms of order.
It seems almost like meeting with some kind of "logical unpredictability", which nevertheless proves possible to measure, even if at times only approximately. To speak of "logical unpredictability" may seem almost a contradiction in terms, but as has been amply demonstrated by a series of experiments, among all the possible trajectories of disorder, nature only favours certain ones.