depth first search

I just finished The Structure of Scientific Revolutions. One continuing difficulty in describing science is the need to properly characterize the act of intuition and discovery. Kuhn provides a masterful account of the historical and sociological aspects  of scientific revolutions, but his identification of the core generative act of knowledge creation is necessarily enclosed in the vague confines of a blurry black box.

Viewed from a distance, the idea of a paradigm, paradigm-shift, and the more general and complete scientific revolution all depend in part on never really knowing the mechanism that, deep down, really drives the process. Kuhn resorts to all kinds of tricks to avoid having to really tackle with what originates revolutions, instead describing in great and useful detail the pre and post effect periods, and in many ways describing everything that the black box does without ever getting inside to see how the black box does it.

But I guess this leaves the whole scaffolding subject to a kind of attack based less on science and more on the idea of revolution. It is a loaded term, and one that, for certain theories of creativity and intuition is probably misapplied. We can mount various kinds of hypothetical attacks on shifting scientific commitments by pointing out that, had we an actually theory of creativity mechanising the theory generating aspect of the structure, the larger (theory generation + paradigm shift system) might end up looking less then revolutionary.

I did come across an interesting critique of Popper’s notion of falsifiability. The critique reminded me of this famous quote from George Box:

Essentially, all models are wrong, but some are useful.

The problem with falsifiability is that most theories are actually not just falsifiable but are in fact false. What one then needs is a notion of less false, which is precisely analogous to the verification-lite notion  of more correct.

•