Unwelcome ideas - part XIV

P K Feyerabend - From Wikipedia

Paul Karl Feyerabend started out studying physics but switched to philosophy. He became known for his radical views on the philosophy of science where at least he knew something of what he was talking, especially physics. He was particularly interested in how scientific discoveries really occur, rather than the cleaned up narrative we are presented with later. His book Against Method is a fascinating read.

Feyerabend was a real maverick, an out and out relativist. He even described himself as an anarchist, although he became wary of the term because of its political implications. I first came across him via an essay he published as part of a collection by various philosophers. His essay was titled How to be a good empiricist - easily the best of the compilation in my view.

In his essay, Feyerabend's main point was the importance he attached to having at least two competing theories to explain any phenomenon. It doesn’t sound as radical as it is, but many scientists seem to actively oppose this way of working – they tend to look for consensus around one theory, as we do in daily life. In particular, scientists expect new theories to have a considerable degree of consistency with older theories and for the meaning of technical terms to remain unaltered. Feyerabend called these artificial limitations the consistency condition and meaning invariance.

For example, we expect mass to stay as mass, energy to stay as energy and so on. So older theories tend to frame the debate about new theories, with obvious limiting consequences. Yet as Feyerabend points out, for the theory of relativity, mass had to change from a Newtonian property of bodies to a relativistic property of the frame of reference. This was a change in the meaning of mass which few scientists would have contemplated because they would not have expected a new theory to amend such an apparently fundamental concept as mass. A few did though and relativity became a successful theory.

So Feyerabend’s point was a simple one really. Theories are best challenged by other theories, not just by the gathering of data and experimental tests. For example, data may only be recognized as relevant within the framework of existing theories. There may be useful data out there which is unused or even invisible to us because we are blinkered by our existing theories. So to challenge a current theory effectively, we must have some other way of looking at the same data - which means we always need another theory.

Of course theories do turn out to be inconsistent with new data and they do fail an experiment, but they are often adaptable enough to be modified to fit the new data or the experiment. Computer models are the obvious and most notorious example. So what should we do if we don't have an alternative theory? Is it right to modify existing theories or tweak our computer models? Are we improving our theories and models or are we merely preserving our investment in them?

Consider the Large Hadron Collider as an example. Presumably it was designed, among other things, to examine the Standard Model of particle physics. Was it designed to look for alternatives to the Standard Model? I doubt it. So was it worth building? I can't tell, but I'm sure the question is worth asking. Feyerabend would have asked - he was like that.

Feyerabend’s remedy for doctrinal science was to promote competing theories as a matter of good scientific practice. Yes we need as much data as possible and we need to test theories by experiment, but another essential tool is to have at least one alternative and incompatible theory.

So do mainstream climate scientists read Feyerabend? Obviously not or they would presumably not be so stupid as to reject so many worthwhile climate theories in favour of their policy-driven CO2 theory.

As far I can see, there is no reason to restrict this way of thinking to science. Politics and economics for example: we should be open to alternative explanations if we wish to avoid the sterility of doctrinal thinking and the likely futility of applying stock, party-political solutions to social and economic problems.