David McCusker picked up the Feyerabend thread again. Interesting that Dave zeroed in on the Feyerabend quote that ecapsulates the central message of Against Method. That central message is a rebuttal to Popper, Lakatos and others who argued that science proceeds according to some set of methodological rules. Why would anyone argue that case ? Well, firstly there's the so called 'scientific method', often loudly cited by anyone arguing for the primacy of scientific knowledge over other forms of knowledge we may have from literature or religion, for instance. And secondly, as human beings, we crave certainty. If science is delivering certain knowledge, then it's natural to ask how that knowledge is certain. Methodology, like Popper's falsificationism, intends to explain how it is that science delivers.

Anyone see a parallel with software methodology ? Peter Naur did. In Computing: A human activity there's a marvellous paper: Programming as Theory Building in which he explicitly draws the parallel between the construction of scientific theories and software systems.